park kyoungweon 200612 phd

8/12/2019 Park Kyoungweon 200612 Phd

http://slidepdf.com/reader/full/park-kyoungweon-200612-phd 1/241

Synthesis, Characterization, and Self –Assembly of Size Tunable Gold Nanorods

A DissertationPresented to

The Academic Faculty

by

Kyoungweon Park

In Partial Fulfillment

of the Requirements for the Degree

Doctor of Philosophy in the

School of Polymer, Textile and Fiber Engineering

Georgia Institute of Technology

December 2006



Synthesis, Characterization, and Self –Assembly of Size Tunable Gold Nanorods

Approved by:

Dr. Mohan Srinivasarao, AdvisorSchool of Polymer, Textile & Fiber

Engineering


Dr. Robert DicksonSchool of Chemistry and Biochemistry


Dr. Mostafa A.El-SayedSchool of Chemistry and Biochemistry


Dr. Anselm GriffinSchool of Polymer, Textile & Fiber

Engineering


Dr. Jung O. Park

School of Polymer, Textile and FiberEngineering


Dr. Laren M. Tolbert

School of Chemistry and Biochemistry


Date Approved: September 19, 2006



iii

ACKNOWLEDGEMENTS

To my research advisor, Dr. Mohan Srinivasarao, for his support, guidance, and

encouragement.

To my committee, comprised of Drs. Jung O. Park, Anselm Griffin, Mostafa El-Sayed,

Robert Dickson, and Laren Tolbert, for their guidance and cooperation during my

research.

To my research group, Jong Seung Park, Vivek Sharma, Matije Crne, Lulu Song , Jian

Zhou, Haixia Wu and Raymond René, for great discussions and enjoyable moments.

To the members of the New Church of Marietta, for their love and prayer.

To my parents, Jae-Yoon Park and Sam-Sun Hur, and my sisters, Kyoung-Hee, Kyoung-

Ah and my brother Mum-hm, for their bottomless love, support, and prayer. I could

not have done this without them.

Most especially, to my son, Caleb for making my life full of delight which I have been

blessed with since he was born.



iv

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ..........................................................................................iii

LIST OF TABLES .........................................................................................................ix

LIST OF FIGURES..........................................................................................................x

LIST OF SCHEMES ....................................................................................................xxii

SUMMARY.................................................................................................................xxiii

Chapter I Introduction ...................................................................................................1

References ............................................................................................................6

Chapter II Synthesis of Refined Gold Nanorods in Binary Surfactant System.............8

2.1. Introduction......................................................................................................9

2.1.1. The effect of surfactants.......................................................................12

2.1.2. The reduction .......................................................................................14

2.2. Experimental..................................................................................................17

2.2.1. Preparation of the seed solution...........................................................17

2.2.2. Preparation of the growth solution.......................................................18

2.2.3. Instrumentations...................................................................................20

2.3. Results and Discussion ..................................................................................20

2.3.1. The role of CTAB in enhancing the formation of anisotropic shape of

gold nanorods. ....................................................................................................20



v

2.3.2. The role of BDAC in enhancing the formation of anisotropic shape of

gold nanorods. ....................................................................................................31

2.3.3. The reduction .......................................................................................39

2.3.4. Tuning aspect ratio of NRs by kinetically controlled growth..............51

2.3.5. The effectiveness of seed-mediated method ........................................53

2.4. Conclusions....................................................................................................55

References ..........................................................................................................56

Chapter III Morphology and Growth mechanism of Gold Nanorods .........................60

3.1. Introduction....................................................................................................61

3.2. The crystallographic model of gold nanorods ...............................................65

3.3. Growth mechanism ........................................................................................70

3.4. Conclusions....................................................................................................76

References ..........................................................................................................77

Chapter IV Shape and Size Separation Gold Nanorods ..............................................78

4.1. Introduction....................................................................................................79

4.2. Theory............................................................................................................82

4.3. Experimental..................................................................................................86

4.3.1. Centrifugation of as made solution......................................................86

4.3.2. Effect of surfactant...............................................................................87

4.3.3. Precipitation of nanorods in gravitational field. ..................................87

4.4. Results and Discussion ..................................................................................88

4.4.1. Separation of nanorods from spherical particles..................................88



vi

4.4.2. Separation of different aspect ratios. ...................................................93

4.4.3. Effect of surfactant concentration........................................................94

4.4.4. Separation by flocculation of rods in gravitational field .....................99

4.5. Conclusions..................................................................................................102

References ........................................................................................................104

Chapter V The Optical Properties of Gold Nanorods................................................107

5.1. Introduction..................................................................................................108

5.2. Experimental ................................................................................................ 113

5.3. Results and Discussion ................................................................................115

5.3.1. Color of gold nanorods in aqueous solution. .....................................115

5.3.2. Effect of refractive index of medium.................................................121

5.4. Conclusions..................................................................................................128

References ........................................................................................................129

Chapter VI Optical Properties of PVA/Gold Nanorods nanocomposites..................132

6.1. Introduction..................................................................................................133

6.2. Experimental................................................................................................137

6.3. Results and discussion .................................................................................138

6.3.1. Optical properties of gold NRs in PVA..............................................138

6.3.2. Polarization dependent color filters ...................................................141

6.4. Conclusions..................................................................................................149

References ........................................................................................................150



vii

Chapter VII Pattern Formation of Gold Nanorods by Drying-Mediated Self Assembly

......................................................................................................................................152

7.1. Introduction..................................................................................................153

7.2. Experimental................................................................................................158


7.3.1. Interparticle Forces among nanorods.................................................160

7.3.2. Heterogeneity and polydispersity of the sample................................161

7.3.3. Patterns formed by homogeneous evaporation ..................................165

7.3.4. Patterns formed by heterogeneous evaporation .................................169

7.3.5. Self assembly formed at the edge of the evaporating drop................176

7.3.6. Effect of molecular linkage on the self assembly of gold NRs..........179

7.4. Conclusions..................................................................................................180

References ........................................................................................................182

Chapter VIII Lyotropic Liquid Crystalline Phases of Colloidal Gold Nanorods

Observed in Evaporating Drop .....................................................................................185

8.1. Introduction..................................................................................................187

8.1.1. The LC phase of colloidal suspension ...............................................185

8.1.2. Contact line deposits in an evaporating drop.....................................190

8.2. Experimental................................................................................................192


8.3.1. The “coffee ring” of gold nanoparticle suspension............................193

8.3.2. LC phase formed in an evaporating drop...........................................195

8.3.3. Two dimensional phase transition observed in self assembly on a



viii

TEM ...........................................................................................................200

8.3.4. Patterns formed by LC in evaporating solution. ................................203

8.4. Conclusions..................................................................................................208

References ........................................................................................................210

Chapter IX Conclusions and Recommendations for the Future Work ......................213

9.1. Conclusions..................................................................................................213

9.2. Recommendations for future works.............................................................215

References ........................................................................................................217



ix

LIST OF TABLES

Table 2.1. Standard potentials in aqueous solutions .......................................................25

Table 2.2. Maximum LSP(λmax), length, diameter and aspect ratio of gold NRs

synthesized at different temperature............................................................................. 51

Table 2.3. The modified reduction steps to obtain higher aspect ratio of gold NRs. .....52

Table 3.1. Particle size analysis of gold NRs using different sampling area and number

of sample. .......................................................................................................................67

Table 4.1. The size of NRs before and after the separation (unit:nm)............................92

Table 5.1. The refractive index of water/glycerol mixture. ..........................................124



x

LIST OF FIGURES

Figure 2.1. Influence of the type of surfactant used in growth solution on the

morphology of the gold NRs: a) CTAC and b) CTAB . Scale bar is 50nm....................21

Figure 2.2. The spectral change in aqueous solutions of HAuCl4 upon adding different

surfactants. [HAuCl4] = 5×10-4 M for all solutions and [surfactant]= 5×10-2 M if added

a) HAuCl4, b) HAuBr 4, c) HAuCl4/CTAB, d) HAuCl4/CTAC, and e)

HAuCl4/KBr(0.05M) mixture.........................................................................................23

Figure 2.3. The spectral change in aqueous solutions of HAuCl4 upon adding increasing

amount of 0.1M of KBr solution. [HAuCl4] = 5×10-4 M for all solutions and [CTAC]=

5×10-2 M, [KBr]= a: 0, b: 0.003M, c: 0.004M, d: 0.005M, e: 0.006M , f: 0.008M, and

g: 0.035M. As the concentration of KBr increased, the characteristic peak of 220 and

300nm shifted gradually to 260 and 380nm. ..................................................................24

Figure 2.4. UV-Vis spectrum of gold NR solution. Effect of KBr on the formation of

NR in the growth solution containing CTAC. ................................................................26

Figure 2.5. A TEM image showing that significant amount of short rods are obtained

when KBr is added with CTAC. Scale bar is 50nm. ....................................................27

Figure 2.6. UV-Vis-NIR spectra of gold NR solution prepared by different gold salts. 28

Figure 2.7. AFM images of adsorbed surfactant layers on silica observed at 10×CMC29 . a) CTAB and b) CTAC..............................................................................................29

Figure 2.8. The effect of CTAB concentration on the formation of gold NRs. a:

[CTAB]=0.02M, b: 0.05M, c: 0.08M, d: 0.1M........................................................30

Figure 2.9. TEM images of gold NRs synthesized in a) CTAB and b) CTAB/BDAC.

Scale bar is 50nm............................................................................................................31

Figure 2.10. The spectral change in aqueous solutions of HAuCl4 upon adding

different surfactant. [HAuCl4] = 5×10-4 M for all solutions and [surfactant] = 5×10-2

M.....................................................................................................................................32

Figure 2.11. Change of the absorbance of the surfactant-HAuCl4 mixture solution at

600nm as a function of total surfactant concentration. The concentration of HAuCl4 was



xi

fixed at 5×10-4 M. The absorbance at 600nm is not related to any of the characteristic

peak of the solutions and is an indicative of the turbidity. .............................................33

Figure 2.12. UV-Vis-NIR spectra of gold NR solutions having different total

concentration of surfactant. The ratio of CTAB/ BDAC was 0.5. The total

concentration of surfactants was a) 0.135M, b) 0.180M, c) 0.225M, and d) 0.270M....34

Figure 2.13. UV-Vis spectra of the gold NRs solutions with increasing ratio of

CTAB/BDAC when the concentration of BDAC is fixed at 0.125M. ...........................35

Figure 2.14. TEM images of gold NRs with the ratio of CTAB/ BDAC a) 0.2, b) 0.4

c) 0.6, and d) 1. The concentration of BDAC is fixed at 0.125M. The scale bar is

100nm.............................................................................................................................35

Figure 2.15. The effect of the ratio of BDAC/CTAB on the size of NRs. The

concentration of BDAC is fixed at 0.125M. It can be seen that as the CTAB fraction

increases, the length of NRs increases and the diameter of NRs decreases at first and

increases continuously. ...................................................................................................36

Figure 2.16. The effect of the ratio of CTAB/BDAC on the growth solutions. The


increases, the length of NRs increases and the reproducibility increases. .....................36

Figure 2.17. UV-Vis-NIR spectra of the gold NRs solutions with decreasing ratio of

BDAC/CTAB when the concentration of CTAB is fixed at 0.1M. The concentration of

BDAC is a) 0.03M b) 0.075M, and c)0.125M. ..............................................................37

Figure 2.18. The effect of reducing agent on the formation of gold NRs. .....................40

Figure 2.19. UV-Vis-NIR spectra of 6 identical growth solutions in which the ascorbic

acid(AA) contents increased from a to f. The ratio of AA:gold ion was a) 1.1, b)1.2 , c)

1.3, d)1.4, e) 1.5, and f) 1.6. The spectra were taken from the as-made solutions after

the growth was ended (measured in 1mm path quartz cell). ..........................................42

Figure 2.20.Changing morphology of NRs by varying the molar ratio of AA to Au +3.

a: 1.1, b: 1.3, and c: 1.6 (a′, b′, and c′ show the byproduct from the same solution).

Scale bar is 50 nm...........................................................................................................43

Figure 2.21. Effect of ascorbic acid(AA) on the growth kinetics of gold NRs. The

molar ratio of AA to gold ion was changed from 1.1 to 1.9. The intensity at LSP was

taken and normalized by the final intensity. ...................................................................44



xii

Figure 2.22. Effect of ascorbic acid(AA) on the size of NRs and nanocubes(NCs). The

lenght of the NRs first increases and then decreases with increasing amount of AA. The

diameter of NRs and the size of NCs increases as the amount of AA increases. ...........44

Figure 2.23. UV-Vis-NIR spectra of the NR solutions before and after adding extra

ascorbic acid (AA) to the mother solution which was initially made with the molar ratio

of AA to Au+3 is 1.1 (a: mother solution, b:adding 10µl, c:30µl, d: 50µl, and e:

100µl). ............................................................................................................................46

Figure 2.24. TEM images of gold NRs and the size distribution of each sample; a)

from batch solution, b) grown from the batch solution (10ml) by adding 30 ul of extra

AA, and c) grown from the batch solution (10ml) by adding 50 ul of extra AA. Scale

bar is 50nm. ....................................................................................................................47

Figure 2.25. The growth of NRs by reinitiating the reducing process. The length of the

NRs was increased with increasing amount of extra AA added after the first growth was

completed. The width in the middle didn’t change significantly, while the width in the

end increases as the amount of AA increases. ................................................................48

Figure 2.26.TEM images of gold NRs grown from mother solution by adding extra

amont of AA. Bone shaped NPs as well as boomerang shaped NPs were obtained

when the largest amount of AA was added.....................................................................48

Figure 2.27. The different morphology of gold NRs. a) 100ul of 0.1M AA was added

initially, and b) 70ul of 0.1M AA was added in the beginning and 30ul was added after

24 hrs. Scale bar is 50nm................................................................................................49

Figure 2.28. Changing morphology of NRs by varying the reaction temperature. a)

50°C, b) 30°C and c) 20°C. All other experimental parameters are the same. The scale

bar is 100nm. ..................................................................................................................50

Figure 2.29. TEM images of (a) gold NRs from mother solution and (b) gold NRs

grown by adding small amount of ascorbic acid step by step. The scale bar is 100nm.

........................................................................................................................................52

Figure 2.30. UV-Vis-NIR spectra of the NR solution grown without seed solution. A

very small amount of NaBH4 solution equivalent to the amount of NaBH4 in the seed

solution was added instead of adding the seed solution. In 24hrs, only short NRs were

produced and the yield was very low (a). After 1 month , they grew longer and the yield

became higher (b). ..........................................................................................................53



xiii

Figure 2.31. TEM image of gold NRs grown without seed solution. The scale bar is

100nm.............................................................................................................................54

Figure 3.1 The structural model of gold NRs proposed by Wang et al. .......................62

Figure 3.2. The structural model of gold NRs proposed by Gai et al.............................63

Figure 3.3 The TEM images of gold NRs taken from the different field of view of a

specimen, showing different aspect ratios......................................................................66

Figure 3.4. The collection of different shape and size of NPs produced by the one step

seed mediated synthesis in binary surfactant system. a) aspect ratio of 3, b) aspect ratio

of 5, and c) aspect ratio of 7. ........................................................................................67

Figure 3.5. The electron diffraction pattern of gold NRs from a single rod. a) L/d= 24

and b) L/d= 4. Scale bar is 50nm....................................................................................68

Figure 3.6. TEM image of NRs along with bigger MTP and cubic as byproducts. .......68

Figure 3.7. Octagonal cross section of the NRs packed as a square packing.................69

Figure 3.8 a) Possible growth mechanism of gold nanocube. b) Resulting gold cubic

particles. Scale bar is 50nm. ...........................................................................................71

Figure 3.9. TEM image of NRs synthesized in one batch. Scale bar is 50nm. ............72

Figure 3.10. a) The relation between diameter and length of NRs. b) The relation

between aspect ratio and volume of NRs. ......................................................................72

Figure 3.11. TEM image of gold NRs grown without seed solution. Scale bar is 100 nm.

........................................................................................................................................73

Figure 3.12. The relation between length and diameter of NRs. Open circle

represents the NRs prepared without seed and filled square represents the NRs prepared

with seed. The diameter of NRs without seed is smaller when the length is similar. ..74



xiv

Figure 4.1. The ratio of the sedimentation coefficients between rod and sphere as a

function of the ratio of diameters between them. The higher the ratio is, the higher the

chance for rod sediment faster........................................................................................86

Figure 4.2. TEM image of the nanoparticles in the mother(as-made) solution, showing

coexistence of the nanorods and nanospheres. L/d of gold NRs is 7.3. .........................88

Figure 4.3. UV-Vis –NIR spectrum of the mother solution. L/d of the gold NRs is 7.3.

........................................................................................................................................89

Figure 4.4. a) Schematic drawing of a centrifuge tube after the centrifugation and the

color of resulting solutions. b) the color of the solution taken from two different

locations shown in (a).....................................................................................................90

Figure 4.5. UV–Vis-NIR spectra of the solutions of the deposits on the side wall of the

tube and at the bottom. ...................................................................................................91

Figure 4.6. TEM images of gold nanoparticles: a) mother solution, b) after

centrifugation, NRs deposited on the side wall of the tube, and c) after centrifugation,

sedimented at the bottom, nanocubes, spheres and NRs with larger diameter...............92

Figure 4.7. The UV-Vis-NIR spectra of the centrifuged samples as a function of time. 94

Figure 4.8. The plot of relationship between diameter and length of the NRs...............94

Figure 4.9. UV–Vis-NIR spectra of the separated solutions. a) deposit on the bottom

and b) deposit on the side wall as a function of CTAB concentration. ..........................96

Figure 4.10. The yield of obtaining NRs as a function of amount of CTAB contained

in the solution .................................................................................................................97

Figure 4.11. Gold NRs flocculation observed under the microscope with crossed

polarizers. The scale bar is 100µm. ..............................................................................100

Figure 4.12. UV–Vis-NIR spectrum of separated solutions.......................................101



xv

Figure 5.1. Absorbance spectra calculated with the expressions of Gans for elongated

ellipsoids using the bulk optical data for gold. a) The numbers on the spectral curves

indicate the aspect ratio ( L/d ). b) Enlargement of the shaded area of a) showing slight blue shift of TSP on increasing aspect ratio. ................................................................ 111

Figure 5.2. Longitudinal surface plasmon peak versus the aspect ratio of gold NRs.

Simulation results using the DDA method and the corresponding fit (solid line) and

Gans’ calculation (dotted line). Experimental data from the work of van der Zande et

al(square). Experimental data from our study (triangle). .............................................117

Figure 5.3. The simulated color of gold NR solution of different aspect ratios and their

trichromatic coefficients. ..............................................................................................118

Figure 5.4. a) UV-Vis-NIR spectra of gold NR solutions having different aspect ratios

and b) transverse peak in detail. .................................................................................119

Figure 5.5. a) Photograph of 4 solutions of colloidal gold prepared in water. Aspect

ratios are 2.6, 4.1, 5.6, 7.4 (from the left), respectively. b) The simulated color of gold

NR solution of different aspect ratios...........................................................................120

Figure 5.6. The color of gold NR solution of having different amount of spheres as

byproduct. a) 50%, b) 30%, c) 10%, and d) 0%.........................................................121

Figure 5.7. Calculated LSP as a function of refractive index of medium. ...................122

Figure 5.8. Absorption spectra of gold NRs with aspect ratio L/d = 3.5 in various

glycerol/ water ratios. From left, water, 20, 40 and 50 % (v/v) of aqueous glycerol

solutions........................................................................................................................123

Figure 5.9. Experimental and calculated results of LSP of gold NRs as a function of

refractive index of medium (water/glycerol mixture). Filled dots represent the

experimental data and straight line is calculated data. .................................................124


volume fraction of DMSO in water. Filled dots represent the experimental data and

straight line represents the calculated data. ..................................................................125



xvi

Figure 5.11. Uv-Vis-NIR spectra of gold NRs in water and water acetonitrile mixture

(water: acetonitrile= 1:4). .............................................................................................127

Figure 6.1. a) UV-vis spectra of drawn nanocomposites comprising of high-density

polyethylene and gold, taken in linearly polarized light at different angles between the

polarization direction of the incident light and the drawing direction. b) Twisted-

nematic liquid crystal displays (LCD) equipped with a drawn polyethylene-silver

nanocomposite. The “M” represents the on state, the drawing direction is in the picture

above parallel and below perpendicularly oriented to the polarizer.............................135

Figure 6.2. The absorbance spectra of randomly oriented gold rod dispersions. a) aspect

ratio 4.3 , b) aspect ratio 6.3. The blue line is the absorbance spectra of aqueous gold

NR solution. The red line is the absorbance spectra of gold NRs in PVA film.........139

Figure 6.3. Longitudinal surface plasmon peak (nm) versus refractive index of the

medium. Straight line is obtained from the calculation of Gans theory. Diamonds and

triangles are the experimental data. ..............................................................................140

Figure 6.4. The experimental set-up of the polarization spectroscopy studies. At a

polarization angle θ=0°, the electric field of the light is polarized parallel to the stretch

direction, whereas, at a polarization angle θ=90°, the electric field of the light is

polarized perpendicular to the stretch direction. ..........................................................142

Figure 6.5. The polarization spectra of the gold NRs with L/d=2.8 dispersed in PVA for

various elongation numbers with respect to the original film. Indicated on the spectral

curves. a) θ=0, b) θ=90.................................................................................................142

Figure 6.6. CSLM micrographs of the PVA film containing gold rods with L/d = 7

demonstrating alignment with stretching (a) the unstretched film and (b) stretched

film. ..............................................................................................................................143

Figure 6.7. UV-vis-NIR spectra of PVA/gold NRs nanocomposites for varying

polarization angles. L/d of gold NRs is 2.8. .................................................................144

Figure 6.8. Optical micrographs of drawn PVA-gold nanocomposites with ca 4 wt %

gold NRs and draw ratio 4. Aspect ratio of NR is 2.8. a) unpolarized, b) polarization



xvii

direction parallel to the drawing direction and c) polarization direction perpendicular

to the drawing direction. Scale bar is 50µm.................................................................144

Figure 6.9. Transmittance spectra of PVA/gold NRs nanocomposites for varying

polarization angles. L/d of gold NRs is 2.8. .................................................................145

Figure 6.10. UV-vis-NIR spectra of PVA/gold NPs nanocomposites for varying

polarization angles. Gold NPs consist of NR of L/d= 2.5 and nanosphere of

diamerter=30nm. ..........................................................................................................146


gold NPs mixture and draw ratio 4. a) unpolarized, b) polarization direction parallel to

the drawing direction and c) polarization direction perpendicular to the drawing

direction. Scale bar is 50µm. ........................................................................................146

Figure 6.12. UV-Vis-NIR spectrum of PVA/gold NRs nanocomposites film. The

aspect ratio of NRs is 6.3..............................................................................................147

Figure 6.13. The polarization spectra of the gold NRs with L/d=6.3 dispersed in PVA.

a) Visible region b) near IR region...............................................................................148



direction parallel to the drawing direction and c) polarization direction perpendicular

to the drawing direction. Scale bar is 50µm.................................................................148

Figure 7.1. Self-assembled morphologies resulting from homogeneous evaporation

and wetting of nanoparticle domains. The upper panels show the experimental results.

The lower panels show results of model simulations for coverages of 5% (a), 30% (b),

40% (c) and 60% (d). The upper panels show corresponding experimentally observed

morphologies for CdSe nanocrystals at similar coverages. From reference 5. ............156

Figure 7.2. Self-assembled morphologies resulting from inhomogeneous evaporation in

simulations (left) and in experiments (right). From reference 5...................................157

Figure 7.3. a) A ring-like array formed from the dilute solution of nanocrystals. b) Side-



xviii

view schematic showing a scenario for evaporation of solvent from nanoparticle

solution on wetted substrate (top), and its thinning to thickness thole < te where thole is the

thickness of the film and te is the critical thickness at which point a hole nucleates(middle). The hole then opens, collecting particles in its growing perimeter (bottom),

which becomes an annular ring upon pinning of its contact line. From reference 10..157

Figure 7.4. TEM image of self assembly of as-made solution. (a) Fast evaporation

(condition I) and (b) slow evaporation (condition II). The coverage of the particles in

both images is about 20%.............................................................................................161

Figure 7.5. TEM image of self assembly of as-made solution showing micro phase

separation of 3 different shapes of nanoparticles. ........................................................163

Figure 7.6. The TEM images showing different patterns formed at different coverage

ratio of a) 4%, b) 10%, c) 15% and d) 40%. The aspect ratio is 5 and the scale bar is

1µm...............................................................................................................................166

Figure 7.7. TEM images of self assembly of gold NRs showing transition from ribbon-

like structure to disk-like smectic domain. a) The individual domain of ribbon-like

structure, b) small domain of nematic-like structure, c) domains of smectic-like

structure and d) an enlarged image of c showing an individual domain of smectic-like

structure. .......................................................................................................................167

Figure 7.8. a) TEM images of self assembly of gold NRs showing a large area covered

by monolayer of gold NRs. b) One of the interesting features formed by the

monolayer of gold NRs. The “eye area” is the microphase-separated nanospheres

surrounded by NRs. ......................................................................................................169

Figure 7.9. Ring-like array observed when the coverage is 10 to 20%. a) aspect ratio 4,

b) aspect ratio 6. ...........................................................................................................170

Figure 7.10. The volume of the particle within a ring-like structure decreases toward

outer rim which indicates that lighter particle is pushed further than heavier particle.172

Figure 7.11. The circular domain formed between the ring-like array structures........172



xix

Figure 7.12. TEM images of self assembly of gold NRs showing the formation of the

ring-like structure in different patterns.........................................................................173

Figure 7.13. TEM image of ring array having the monolayer of self assembly as a

background. ..................................................................................................................174

Figure 7.14. TEM image showing the transition from ring-like array to cellular

network structure. .........................................................................................................174

Figure 7.15. The cellular network structures formed by gold NRs with different aspect

ratios: a) aspect ratio of 4 and b) aspect ratio of 6........................................................175

Figure 7.16. Cellular network structure formed in a large area....................................176

Figure 7.17. The self assembly of gold NRs formed at the edge of the evaporating

drop...............................................................................................................................177

Figure 7.18. Different patterns of self assembly formed by evaporating gold NR

solution. ........................................................................................................................178

Figure 7.19. TEM images of gold NRs in the presence of azo dye rotaxane showing

end-to-end linkage mediated by azo dye rotaxane. The aspect ratio of the gold NRs

is 4.5. The concentrations of dye rotaxane increases from a to d. a) 2.0 10-6M, b)

4.08 10-6M, c) 4.9 10-5M and d) 7.35 10-5M. ............................................................180

Figure 8.1. The factors leading to outward flow in a drop of fixed radius R slowly

drying on a solid surface. The evaporative flux J (r ) reduces the height h(r ) at every

point r . As its contact line is pinned, the radius of the drop cannot shrink. To prevent

the shrinkage, liquid must flow outwards v(r). From the reference 16. .......................190

Figure 8.2. Images of a) a 2-cm-diameter drop of coffee from reference 16 and b) a 4-

mm-diameter drop of colloidal gold NRs.....................................................................191

Figure 8.3. The contact angle of a drop of gold NR solution (1µl) on a glass substrate as

a function of time..........................................................................................................194



xx

Figure 8.4. The band width is plotted as a function of normalized volume fraction of the

gold NRs in the case of slow evaporation. ...................................................................195

Figure 8.5. Optical microscope images of a dried drop of gold NRs with the aspect ratio

of 3. a) Bright field image and b) polarized image at the edge. Birefringence is

from the surfactant crystal. ...........................................................................................196

Figure 8.6. In this series of images the center of the drop is in the bottom left corner.

The images of the drop of gold NRs solution during evaporation under optical

microscope with cross polarizers, as described in the text. The scale bar is 30µm......197

Figure 8.7. The images of the drop of gold NRs solution during evaporation under

optical microscope with cross polarizers, as described in the text. a) Emerging liquid

crystal domain from the interior near the edge and moving toward the edge. b)

Magnified image showing the individual domain joins the existing structure. c) The

assembly of liquid crystal domains before complete drying. The scale bar is 200 µm

for a) and 20 µm for b) and c). .....................................................................................197

Figure 8.8. TEM images of a dried drop of gold NR solution. a) The edge of the drop

showing smectic-like structure. b) and c) show the small size of nematic-like domain

away from the edge.......................................................................................................198

Figure 8.9. TEM images of assembly of gold NRs. a) and b) showing the change of

the orientations in assembly of gold. c) and d) showing the identical areas with tilted

angle. The scale bar is 100nm. .....................................................................................199

Figure 8.10. a) Preparation of a gold NR assembly on a TEM grid, b) A polarizing

microscope image showing LC phase uniformly formed over the grid during the

evaporation. ..................................................................................................................200

Figure 8.11. TEM images of gold NR assembly. Aspect ratio is 6. a) Isotropic phase b)

nematic-like assembly, and c) smectic-like assembly. ...............................................201

Figure 8.12. TEM images of gold NRs assembly. Aspect ratio is 3. a) Isotropic phase

and b) transition from isotropic to smectic-like assembly............................................201



xxi


and b) nematic-like assembly. ......................................................................................202

Figure 8.14. Measured values of the gold NRs number density and the order parameter,

S. (a) The aspect ratio is 3 and (b) the aspect ratioj is 6. Number density was

calculated by counting the number of NRs per 10000nm2...........................................202

Figure 8.15. “Coffee stain” formed by drying drops of gold NR solution. The images

in the upper row show the drying drops from slow evaporation. The volume fraction

decreases from left to right: a) 1×10-5 , b) 5×10-6 ,c) 3.3×10-6 d) 2.5×10-6 and f)

1.25×10-6. The images at the bottom row are from fast evaporation. The volume

fraction of the drops in the same column is identical. The scale bar is 200µm............205

Figure 8.16. The patterns formed by various concentrations and different evaporation

condition were visualized by polarized optical microscopy for the samples shown in

Figure 8.15. Only the edge areas are illustrated. The scale bar is 200µm..................205

Figure 8.17. a) A schematic shows the accumulation of gold NRs at the contact line in

the case of slow evaporation. Most of the particles can be carried to the edge forming a

single dark ring. b) In the case of fast evaporation, contact line starts to recede while

the particles move toward the edge resulting in the space between pinning sites........207

Figure 8.18. The well defined concentric multiple ring pattern formed by diluted gold

NR solution. The image is taken under optical microscope with cross polarizers. The

scale bar is 100 µm.......................................................................................................208



xxii

LIST OF SCHEMES

Scheme 2.1. Structures of hexadecyltrimethylammoniumbromide (CTAB) and

benzyldimethylammoniumchloride (BDAC). The latter has a bulkier head group. ...17

Scheme 5.1. Schematic of plasmon oscillation for a metal sphere............................109

Scheme 7.1. A schematic showing the preparation of a TEM grid. ..........................164

Scheme 8.1. Various liquid crystalline phases as a function of concentration of the

solute. .......................................................................................................................186

Scheme 8.2. A schematic of the droplet evaporation process18.................................193



xxiii

Summary

The successful applications of nanoparticles require the ability to tune their

properties by controlling size and shape at the nanoscale. In metal nanomaterial

research, the optical properties have been of interest especially because of the

applications to medical diagnostics and nanooptics. It is important to prepare

nanoparticles of well-defined shape and size for properly characterizing the optical

properties.

We describe improved seed mediated synthesis of gold nanorods (GNRs)

producing a high yield of NRs with low polydispersity and few byproducts. The

efficient separation of GNRs from mixture of shapes is achieved by understanding the

hydrodynamics of nanoparticles undergoing centrifugation. The optical properties of

resulting refined GNRs are compared to predictions of existing theories, and the main

parameters affecting them are discussed.

GNRs with well defined aspect ratios are introduced into a polyvinyl alcohol

matrix by means of solution-casting techniques. The film is drawn to induce the

uniaxial alignment of GNRs to be used as color polarizing filters. We prepare GNR

polarizing filter with different peak positions ranging from visible to near infra red by



xxiv

using different aspect ratio of NRs.

To utilize GNRs to make nanoscale devices, spatial organization is required.

We characterize the self-assembly of GNRs observed on a TEM grid. The drying

process is accompanied by complex hydrodynamic and thermodynamic events, which

create rich range of patterns observed. Being anisotropic in shape, the rods can form

liquid crystal (LC) assemblies above a certain concentration. We observed LC phase of

GNRs by resorting to an evaporation of aqueous NR solution. The convective flow

caused by the solvent evaporation carries NRs from the bulk solution to solid-liquid-

air interface, which makes the solution locally very concentrated driving the phase

transition of NRs. We calculate the order parameter from various assemblies

observed, and compare the observed phase behavior to the one expected on the basis of

theory.



1

Chapter I

Introduction

Gold nanoparticles (NPs) have been known as “colloidal gold” long before the

development of modern gold chemistry. It dates back to the time when colloidal

gold was used as a colorant to make ruby glass and ceramic in the 5th century B.C. 1

However, Faraday was the first to recognize that the brilliant color is due to the

presence of the metallic gold in colloidal form in 18572. He reported the formation

of deep red solutions of colloidal gold by reduction of an aqueous solution of

chloroaurate (AuCl4-) using phosphorus in CS2 (a two-phase system). In the 20th

century, various methods for the preparation of gold colloids were reported 3,4,5.

The most popular one for a long time has been using citrate reduction of HAuCl 4 in

water, which was introduced by Turkevitch in 19513.

With the advent of nanotechnology, gold NPs have turned out to play an

important role in the field because they have the potential to serve as building blocks

for putting nanotechnology to practical use6,7.

Much progress has been made in the synthesis of spherical gold NPs in the

past decades, and the influence of particle radius on the physical, chemical, optical,

electronic and catalytic properties of the material have been broadly studied 8.

By suitable choice of experimental conditions and additives, non-spherical



2

shape of NPs such as rods, wires, tubes and concentric core-shell structures have been

successfully synthesized and they are found to possess properties which depend not

only on the size but also on other topological aspects9. Nanorods (NRs) have drawn

the most attention because they can be synthesized using a relatively simple process

such as wet chemical methods and the rational control over the aspect ratio is possible.

Gold NRs show different color depending on the aspect ratio, which is due to the two

intense surface plasmon resonance peaks (longitudinal surface plasmon peak and

transverse surface plasmon peak corresponding to the oscillation of the free electrons

along and perpendicular to the long axis of the rods)9. The color change provides the

opportunity to use gold NRs as novel optical applications. There have been many

applications utilizing this intense color and its tunablity10. One of them is in the

field of biological system. NRs bind to specific cells with greater affinity and one

can visualize the conjugated cell using a simple optical microscope due to the

enhanced scattering cross section11. This is how gold NRs are used in molecular

biosensor for the diagnosis of diseases such as cancer.

NRs show enhanced fluorescence over bulk metal and nanospheres, due to the

large enhancement of the longitudinal plasmon resonance12, which will prove to be

beneficial in sensory applications. The increase in the intensity of the surface

plasmon resonance absorption results in an enhancement of the electric field and



3

surface enhanced Raman scattering of molecules adsorbed on gold NRs13.

All theses properties make gold NR a good candidate for future

nanoelectronics, once appropriate techniques allow for the generation of artificial

structures in 2D or 3D. All of these properties have lead to a dramatic growth in the

publications related to gold NRs.

The key issue in the successful applications of NPs is how to organize

individual NP into a well defined macroscopic structure having a particular

configuration which shows interesting new collective physical properties that are

different from those of the NPs as well as the original bulk materials. And the

ability to tune their properties by controlling size, shape and surface modifications of

individual particles is very important.

While the number of publications on gold NR research, to date, has been

increasing dramatically, the specific mechanisms governing the morphology and

geometry control over particle growth are still far from being well understood, and the

reproducibility of the synthesis is poor. Therefore, the application is often

demonstrated without the proper characterization of the sample, thus leading to results

that are not necessarily reproducible by various research groups.



4

In this study, we have to carefully re-examine the process of synthesis,

characterization and application of gold NR to have better understanding and insight

of gold NR research. The objective of this study can be summarized as following.

• To gain a better understanding of growth mechanism of gold NRs.

• To achieve the ability to tune the shape and size of gold NRs, quite reproducibly.

• Proper characterization of optical properties, with well characterized samples,

obtained by understanding the principles behind the shape separation of particles.

• To fabricate superstructures using well defined gold NRs.

Following this introduction, Chapter 2 describes the modified seed-mediated

synthesis of gold NRs by which the morphology, mainly the aspect ratio of gold NRs

can be controlled by the choice of experimental conditions and additives. The

parameters such as the type, concentration and combination of surfactants, reducing

agent, temperarture, and additives are extensively discussed. We also describe the

manipulation of the synthesis steps to obtain a desired aspect ratio of gold NRs. In

chapter 3, growth mechanism of gold NRs will be proposed based on the data in

literature and our experimental results in chapter 2. Chapter 4 describes the

hydrodynamic behavior of NPs of various shapes to identify the parameters to achieve

efficient separation of NRs from other shape of NPs. Chapter 5 and 6 are devoted to



5

the optical properties of gold NRs in various dielectric media including polymer

matrix, and the potential application of polarizing color filter is described. In

chapter 7, we demonstrate and analyze varied range of structured aggregates and

spatial patterns formed by the drying of thin gold NRs solution films. In chapter 8,

we accomplish the experimental observations of lyotropic liquid crystalline phases of

colloidal gold NRs by achieving the phase transition volume fraction through the

convective flow resulting from the solvent evaporation.



6

References

1 Pollard, A.P.; Heron, C. “Archaeological Chemistry” Royal Society of Chemistry,

1996; Chapter 5.

2 Faraday, M. “Experimental Relations of Gold (and other metals) to Light” Philos.

Trans. 1857, 147 , 145.

3 Turkevitch, J.; Stevenson, P. C.; Hillier, J. “Nucleation and Growth Process in the

Synthesis of Colloidal Gold” Discussions of the Faraday Society 1951, 11, 55.

4 Frens, G. “Controlled Nucleation for the Regulation of the Particle Size in

Monodisperse Gold Suspensions. Nature: Phys. Sci. 1973, 241, 20.

5 Hayat, M. A. “Colloidal Gold, Principles, Methods and Applications” Academic

Press: New York, 1989.

6 Pellegrino,T.; Kudera, S.; Liedl,T.; Javier,A.M.; Manna,L.; Parak,W.J. “On the

Development of Colloidal Nanoparticles towards Multifunctional Structures and

their Possible Use for Biological Applications” Small 2005,1,48.

7 Elghanian, R.; Storhoff, J.J.; Mucic, R.C.; Letsinger, R.L.; Mirkin, C.A. “Selective

colorimetric detection of polynucleotides based on the distance-dependent optical

properties of gold nanoparticles” Science 1997, 277 ,1078.

8 Marie-Christine,D.; Didier,A. “Gold nanoparticles: assembly, supramolecular

chemistry, quantum-size-related properties, and applications toward biology,

catalysis, and nanotechnology” Chemical reviews 2004, 104, 293 and

references therein.

9 Kelly, K. L.; Coronado, E.; Zhao, L. L.; Schatz, G. C. “The optical properties of

metal nanoparticles: The influence of size, shape, and dielectric environment” J.

Phys.Chem. B. 2003, 107 , 668.

10 Pérez-Juste,J.; Pastoriza-Santos ,I.; Liz-Marzán ,L.M.; Mulvaney, P. “Gold

nanorods: Synthesis, characterization and applications” Coordination Chemistry

Reviews 2005, 249, 1870.

11 El-Sayed,I.H.; Huang,X.; El-Sayed,M.A., “Surface Plasmon Resonance



7

Scatteringand Absorption of anti-EGFR Antibody Conjugated Gold Nanoparticles in

Cancer Diagnostics: Applications in Oral Cancer” Nano Lett., 2005, 5, 829.

12

Eustis,S.; El-Sayed, M.A., “Aspect Ratio Dependence of the EnhancedFluorescence Intensity of Gold Nanorods: Experimental and Simulation Study” J.

Phys. Chem. B 2005, 109, 16350.

13 Nikoobakht, B.; Wang, J.; El-Sayed, M. A., “Surface-enhanced Raman

scattering of molecules adsorbed on gold nanorods: off-surface plasmon resonance

condition” Chemical Physics Letters 2002, 366 , 17.



8

2. Chapter II

Synthesis of Refined Gold Nanorods in Binary Surfactant System

Abstract

The morphology of gold nanorods(NRs) is controlled by manuplating the

reduction condition thus changing the reduction kinetics and the adsorption kinetics of

the surfactant on the surface of growing NR. The parameters affecting the

adsorption kinetics such as types, and concentration of surfactant can also affect the

reduction kinetics. Hexadecyltrimethylammonium bromide (CTAB) forms the

precusor of gold-surfactant complex which has a lower redox potential to influence

reduction reaction. Benzyldimethylhexadecylammonium chloride (BDAC) as a

cosurfactant lowers the 1st and 2nd critical micelle concentration(CMC) of the

surfactant solution promoting the formation of bigger micelle with anisotropic shape

which is benefitial to keep the precursor inside to prevent instant reduction and the

efficient confinement of growing surface of gold NR. Our results show that the

adsorption competes with the reduction of gold ion on the same surface of NR.

When reduction rate is slower, long NRs with smaller diameter are obtained while

when reduction is faster, short NRs with larger diameter are obtained.



9

2.1. Introduction

A common top down technique to prepare gold nanorods (NRs) is electron

beam lithography 1 . This technique has the advantage of preparing highly

monodisperse NRs but it can only create a 2-dimensional structure in a single step and

the instrumentation is expensive. For bottom up techniques, two different

approaches are available to synthesize gold NRs. One is using hard templates to

physically confine the growth such that gold ions are only reduced inside this

template2. The growth mechanism is straightforward and the size and shape of the

NRs are predetermined by the size and shape of the template. This method has

advantages for obtaining monodisperse particles. But large scale synthesis is limited

since it requires nanoporous templates. In the other method, the gold ions are

reduced in the presence of capping materials (also called stabilizers since it stabilizes

the colloidal NPs) which bind to the surface of growing nanoparticle to direct the

formation of NRs (we will call this method as a single batch preparation). As for

the reduction path, one can choose photochemistry (UV-irradiation), electrochemistry,

bio-reduction or chemical reduction. The size and shape of the NRs are determined

by the various experimental parameters that affect the growth mechanism. However,

the specific mechanisms governing morphology and geometry control over particle

growth are not yet explicitly understood.



10

We have been particularly interested in the single batch preparation of gold

NRs where various experiment parameters such as concentration of reactants and

temperature have an effect on the morphology of NRs.

In the following, we summarize the most commonly used methods to grow

gold NRs by chemical reduction.

Gold nanoparticles (NPs) has bee known as colloidal gold for centuries. This

dates back to the time when colloidal gold was used as a colorant to make ruby glass

and ceramic in the 5th century B.C.3 However, Faraday was the first to recognize in

18574 that the brilliant color is due to the presence of the metallic gold in colloidal

form. He reported the formation of deep red solutions of colloidal gold by

reduction of an aqueous solution of chloroaurate (AuCl4-) using phosphorus in CS2

(a two-phase system). In the 20th century, various methods for the preparation of

gold colloids were reported and reviewed5,6. The most popular one for a long time

has been using citrate reduction of HAuCl4 in water, which was introduced by

Turkevitch in 19517.

Since rod-like gold particles were observed as a byproduct of spherical NPs in

19928, many efforts have been made to move beyond conventional spherical growth

to achieve shape control.

In the first observation of rod-like gold particles, UV-irradiation was used as a



11

reduction path and with alkyltrimethylammonium chlorides as a stabilizer.

Synthesis condition such as type and concentration of surfactant and duration of UV-

irradiation has been modified to prepare NRs in majority9,10,11. However, the results

have been rather unsuccessful. Kim et al. also used UV-irradiation to produce NRs

12. They changed the type of stabilizer to hexadecyltrimethylammonium bromide

(CTAB) and tetradodecylammonium bromide. They were able to prepare NRs as

the major component with the aspect ratio up to 5. The aspect ratio of the rods was

controlled by the amount of silver ions (of AgNO3) added to the system. More

recently, combination of chemical reduction of tetrachloroaurate by ascorbic acid and

subsequent UV-irradiation resulted in the quick generation of gold NRs13. By UV-

irradiation method, gold NRs having aspect ratio less than 5 are obtained. However,

this method has a limitation to produce a large quantity of NRs since the reactor size

is limited.

Meanwhile, an electrochemical method was developed by Yu et al. in which the

gold ions were reduced on a platinum electrode in the presence of a solution of

surfactant mixture14. CTAB and tetraoctylammonium bromide (hydrophobic) were

used as surfactants above their critical micelle concentration (CMC). This method is

simple, and NRs can be prepared in a short time.



12

Jana et al. 15 modified the seed-mediated synthesis of nanospheres16 to obtain

NRs by changing the concentration of surfactant and adding AgNO3. Gold

nanospheres are made by chemical reduction of a gold salt with a strong reducing

agent (such as sodium borohydride). These seeds are then added to a growth

solution of gold salt, a weak reducing agent and CTAB. No metal salts are reduced

to metal unless the seeds are present where the seeds serve as nucleation sites for NRs.

The seed-mediated synthesis of gold NRs provides rods with higher aspect

ratios than those prepared by other methods. This method was further developed

into sequential growth method in which NRs grown from the previous synthesis were

used as seed to grow longer NRs(eg., the aspect ratio of 25)17.

Besides the methods mentioned above, various other approaches have been

attempted to produce gold NRs, such as bio-reduction18 and growth of gold NRs

directly on mica surface19.

In any case, the ultimate aim of the research has been to obtain controlled size

and shape of NRs with less byproduct and to understand the effect of various

experimental parameters on the growth mechanism and finally on the morphology of

NRs.

2.1.1. The effect of surfactants

The role of surfactant in NRs formation is critical since both the growth and



13

the shape of the particles are greatly controlled by the condition of surfactant used.

At first, it was believed that the rod-like micelle formed by surfactant provides soft

template for rod formation13. But recently many researchers concluded that

surfactant monomer adsorption takes place preferentially on specific faces so that the

growth in the diameter direction is constrained promoting the growth in the length

direction20 . In fact, the concentration of CTAB used in most synthesis is lower than

the 2nd CMC above which the micelle becomes rod shape21.

CTAB is widely used in the synthesis of NRs (or NPs) in all reduction paths.

The concentration of CTAB determines the shape of NPs. Below CMC, the

reduction primarily results in nanospheres22. Only above CMC, anisotropic shape

develops as the major product15.

Binary surfactant system was originally used in electrochemical method14 and

later adapted for other reduction paths. Surfactants which have much more

hydrophobic head groups were usually used as cosurfactant to induce the rod shape.

Babak et al. were the first to introduce binary surfactant system in the seed-

mediated method20. In this case, more hydrophobic surfactant

(benzyldimethylammonium chloride (BDAC)) was used along with CTAB. By

increasing the ratio of BDAC/CTAB from 2 to 16, the aspect ratio of NRs increased.

Thus, this approach provides rods with fairly good uniformity, higher yield, and yet



14

little byproduct. However, the role played by the cosurfactant was not clearly

understood.

The role of CTAB in the synthesis of NRs has been extensively discussed

15,17,20 and the consensus is that CTAB adsorbs preferentially on to the specific crystal

facet of the gold NPs, thus confining the growth direction.

2.1.2. The reduction

In most methods of synthesis of gold NPs, the chemical principle is the

reduction of gold (III) derivatives, chloroaurate (AuCl4-) in an aqueous solution.

Reduced gold atoms start to form small sub-nanometer particles (nucleation). The

gold atoms that appear later stick to the existing particles (growth). To prevent the

particles from aggregating, the solution is stirred vigorously and a stabilizing agent is

usually added to the mixture.

In UV-irradiation and electrochemical reduction, Au+3 is reduced to Au0. In

the seed-mediated method, ascorbic acid (AA) is often used as a mild reducing agent

to reduce the gold ion from Au+3 to Au+1, and then the reduction to Au0 only begins

when the seed solution is added. The amount of AA added was not the same in

scattered reports and the variation in the size and shape of the NPs produced was not

trivial 15,20. The effect of the type and concentration of reducing agent on the size

and shape of spherical gold particle was investigated23,24. But little attention has



15

been devoted to the effect of reducing agent on the formation of NRs.

The driving force for the reaction is the difference between the redox

potentials (∆E) of the two half cell reactions (reduction of gold ion and oxidation of

reducing agent). An increase in the value of ∆E translates into a more spontaneous

reaction. The kinetics of reaction is also governed by the concentration of reactants,

temperature and pH. Therefore, by changing the type of reducing agent, pH,

temperature, and the concentration of reactants, one can exert control on the size,

shape and morphology of the particle. As for the spherical particles, it is not so

complicated to relate the parameters that affect the reaction kinetics to the final

morphology. For example, strong reducing agents such as NaBH4 or phosphorus

produce small gold particles while mild reducing agents like ascorbic acid produce

larger gold particle which can be explained by decreasing ∆E. Decreasing the particle

size upon increasing the concentration of the stabilizing agent can be explained by

better stabilization24. A mechanism for anisotropic particle formation is still a

problem awaiting complete solution.

In attempting to provide a mechanism for the synthesis of gold NRs, the key

issue is why the particle grows faster in one direction in comparison to the other

direction. The role of CTAB in the synthesis of NRs has been extensively discussed

15,17,20 and the consensus is that CTAB adsorbs preferentially on to the specific crystal



16

facet of the gold NPs, thus confining the growth direction. Then the adsorption and

reduction kinetics should greatly affect the morphology of NR as has been found in

the literature9-20. Then in order to control the growth of NPs, one should be able to

manipulate the adsorption and reduction kinetics so that the gold particles grow

slowly, while the stabilizing agent can exert control over the growth direction. In

this chapter, we focus on the parameters which we believe exert control over the rate

of reduction and adsorption.

We prepared NRs by the modified seed-mediated method20. As has been noted

above, this method provides rods with high aspect ratios (up to L/d=25), fairly good

uniformity, higher yield, and with fewer byproducts, thus allowing us to focus on

establishing the growth mechanism of NRs. Primitive gold NPs were made by the

chemical reduction of a gold salt with a strong reducing agent such as sodium

borohydride. These seeds are then added to a growth solution that contains gold salt,

a weak reducing agent, surfactants and AgNO3. CTAB was used as stabilizer in both

the seed and the growth solution.



17

2.2. Experimental

CTAB and BDAC (chemical structures are shown in Scheme 2.1) were

purchased from TCI America. The quality of the CTAB and BDAC is critical because

the minor impurities (most likely cetyldimethylamine) can result in much faster rates

of reduction preventing the reproducible preparation of NPs25. The chemicals

purchased from other company do not give reproducible result even after multiple

recrystallization procedures.

All other chemicals were purchased from Sigma-Aldrich. Deionized water (18

MΩ) was used in all the experiments. For preparation of NRs, seed and growth

solutions were made as described below.

Scheme 2.1. Structures of hexadecyltrimethylammoniumbromide (CTAB) and

benzyldimethylammoniumchloride (BDAC). The latter has a bulkier head group.

2.2.1. Preparation of the seed solution

Seed solution was prepared by mixing CTAB solution (5.0 mL, 0.20 M) with

5.0 mL of 0.00050 M HAuCl4. To the stirred solution, 0.60 mL of ice-cold 0.010 M

NaBH4 was added, which resulted in the formation of a brownish yellow solution.

CTAB BDAC



18

Vigorous stirring of the seed solution was continued for 2 min, then the solution was

stirred, it was kept at 25 °C without further stirring.

2.2.2. Preparation of the growth solution

2.2.2.1. Variation of surfactant

The growth solution was prepared by mixing 5 mL of the solution at a

known concentration of surfactant (CTAB, hexadecyltrimethylammonium chloride

(CTAC), BDAC) to 5 mL of 0.001M HAuCl4 solution. 200 µl of 0.0040 M AgNO3

solution was added to the solution at 25 °C. After gentle mixing of the solution,

60µl of 0.10M ascorbic acid was added to the test tube. The color of the growth

solution changes from dark yellow to colorless. The final step was the addition of 10

µl of the seed solution to the growth solution. The color of the solution changes over

the period of time depending on the final size of the NRs.

2.2.2.2. Variation of the amount of ascorbic acid

Seven test tubes of growth solutions were prepared. Each tube contained 10

mL of an aqueous solution of 0.10 M CTAB and 0.495 mg/mL of BDAC. 200 µl of

0.0040 M AgNO3 solution was added to this solution at 25 °C. To this solution, 50 µl

of 0.10 M HAuCl4 was added, and after gentle mixing of the solution 50, 55, 60, 65,

70, 75 and 80µl (corresponding ratio of the molar concentration of the reducing



19

agent to the concentration of the gold cations is 1.0, 1.1, 1.2, 1.3, 1.4, 1.5 and 1.6,

repectively) of 0.10M ascorbic acid was added to each tube. All seven growth

solutions above were identical except for their AA content. The color of the growth

solution changes from dark yellow to colorless. The final step was the addition of 10

µl of the seed solution to the growth solution. The temperature was kept at 25°C

during the growth process. The growth rate was found to level off within about 12

hrs.

For reinitiating the growth, 50ml of bulk NPs solution was made as described

above with AA/Au molar ratio of 1.1. After 48 hrs, the bulk solution was divided

into 5 tubes each containing 10ml of NPs solution. 10,30,50,100 µl of 0.10M of

ascorbic acid were added to each tube, which makes the final molar ratio of AA to Au

as 1.3, 1.7, 2.1, and 3.1. Additional ascorbic acid reinitiates the growth within a few

minutes as detected by the change of UV-Vis-NIR spectrum and the growth rate

levels off within 3 hrs.

All the resulting solutions were centrifuged at 5600G for 30 min to remove the

surfactants. The clear supernatant liquid which mainly contains surfactants was

removed, then the concentrated NP solution was collected and redispersed in

deionized water. This process was repeated for further purification. No process for



20

separating shapes and size was done. The process for separating the shapes and sizes

is decribed in chapter 4.

2.2.3. Instrumentations

UV-Vis-NIR spectra were acquired with a Cary 5G UV-Vis-NIR

spectrophotometer. Morphology and the mean size of NPs were examined by TEM

(JEOL100 at 100 KV). For each sample, the size of more than 300 particles were

measured to obtain the average size and the size distribution.

2.3. Results and Discussion

2.3.1.

The role of CTAB in enhancing the formation of anisotropic shape of goldNRs.

Alkyltrimethylammonium surfactant is a prerequisite as a stabilizer for the

formation of anisotropic shape of gold NPs for all reduction paths. CTAB has been

shown to be most effective to produce longer NRs in a larger fraction among other

alkyltrimethylammonium bromide surfactants having different hydrocarbon tail26.

Hexadecyltrimethylammonium chloride (CTAC) was used only for the UV-

irradiation method only to produce NRs in minority8. Since CTAC has never been

used as a single component in other reduction paths, we want to probe how the

difference of counter ion affects the morphology of resulting NRs.



21

When only CTAC was used, only spherical particles were obtained. Short

NRs were obtained in majority with CTAB (Figure 2.1).

Figure 2.1. Influence of the type of surfactant used in growth solution on the

morphology of the gold NRs: a) CTAC and b) CTAB . Scale bar is 50nm.

Since CTAB and CTAC have the same hydrophobic group, the difference in

the morphology of gold NRs can be attributed to different counter ions, Br and Cl

respectively. The observation brings the question, “how does Br ion promote the

growth of NRs over Cl ion?”.

Transition metal complexes such as [PdCl4]-2 and [AuCl4]

- show specific

interactions with oppositely charged surfactants 27. The complex of [AuCl4] – and

CTAC was studied8. Upon mixing HAuCl4 solution to CTAC solution, brilliant

yellow-white fine crystals of gold surfactant complex were produced. The binding

originates from electrostatic attractions between the complex anions and the organic

cations and the existence of 1:1 stoichiometry of CTAC and [AuCl4]- has been

reported8. Above CMC of CTAC, the complex is solubilized and the solution

became clear.



22

Our visual observations during the preparation of gold NRs also indicated that

[AuCl4]- interacts with CTAB before the reduction process is induced. But the

change of color is more pronounced than that of CTAC and [AuCl4]- mixture.

Upon addition of CTAB to the pale-yellow solution of HAuCl4, orange color

precipitate forms. As the concentration of CTAB increases, the precipitation

disappears and the color of the solution becomes orange. In the later stage, water-

insoluble gold-surfactant complex is solubilized into micelles.

The UV-Vis spectrum of HAuCl4 solution has two peaks at 220 and 300 nm

due to the ligand metal charge transfer 28. By mixing HAuCl4 solution with different

surfactant, the peak position is shifted depending on the interaction between the

surfactant and [AuCl4]-. Therefore the complex can be characterized by the peak

shift. When HAuCl4 solution is mixed with CTAB solution, the ligand metal

charge transfer peaks are shifted to 260 and 380nm which are very close to the peaks

of [AuBr 4]- as compared in the Figure 2.2. The change is caused by the change of

ligand X in [AuX4]- from Cl to Br, since it is easier to oxidize Br than Cl 28.



23

200 300 400 500

0.0

0.5

1.0

1.5

Wavelength (nm)

0.0

0.5

1.0

1.5

0.0

0.5

1.0

1.5

0.0

0.5

1.0

1.5

e

d

c

a

b

0.0

0.5

1.0

A b s o r b a n c e

( a . u .

)

A b s o r b a n c e

( a . u .

)

A b s o r b a n c e

( a . u .

)

A

b s o r b a n c e

( a . u .

)

A b s o r b a n c e

( a . u . )

Figure 2.2. The spectral change in aqueous solutions of HAuCl4 upon adding different

surfactants. [HAuCl4] = 5×10-4 M for all solutions and [surfactant]= 5×10-2 M if

added. a) HAuCl4, b) HAuBr 4, c) HAuCl4/CTAB, d) HAuCl4/CTAC, and e)

HAuCl4/KBr(0.05M) mixture.



24

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

200 300 400 500 600 700

Wavelength(nm)

O p t i c a l d e n s i t y

a → g


increasing amount of 0.1M of KBr solution. [HAuCl4] = 5×10-4 M for all solutions

and [CTAC]= 5×10-2 M, [KBr]= a: 0, b: 0.003M, c: 0.004M, d: 0.005M, e: 0.006M ,

f: 0.008M, and g: 0.035M. As the concentration of KBr increased, the characteristic

peak of 220 and 300nm shifted gradually to 260 and 380nm.

For comparison, we added KBr to the solution of HAuCl4 and examined the

spectral changes. As the concentration of KBr increased, the characteristic peaks at

220 and 300nm shifted gradually to 260 and 380nm (Figure 2.3).

Therefore the net interaction between CTAB and HAuCl4 in aqueous solution

results in the formation of [CTA]-[AuBr 4], while CTAC and [AuCl4]- form a

complex without the change of ligand. This organic salt is solubilized by the

surfactant micelles to produce “metallomicelles”.

+−− ++][↔+] CTA3Cl4AuBr -CTACTAB4[AuCl 44

So, one is left to wonder, “why is it important to have [AuBr 4]-, instead of



25

[AuCl4] – , and metallomicelles regarding the formation of NRs?” If metallic species

are involved in the formation of solute complexes or compounds, the standard redox

potential will be lower since the complex is more stable than the gold ion. In Table

2.1, the standard potentials of gold ion and gold complex are shown29.

Table 2.1. Standard potentials in aqueous solutions29

Half reaction Standard potential(V)

Au(III)/Au(I)

Au3+ + 2e- → Au1+ +1.40

AuCl4- +2 e- →AuCl2

- + 2Cl- +0.926

AuBr 4- + 2e- → AuBr 2

- + 2Br - +0.805

Au(I)/Au(0)

Au+ + e- → Au +1.71

AuCl2- + e- →Au + 2Cl- +1.154

AuBr 2- + e- → Au + 2Br - +0.962

Ascorbic acid

C6H6O6+ 2 H+ + 2 e- → C6H8O6 +0.13

The potential of gold complex is lower than that of gold ion. On top of it,

[AuBr 4]- has a lower potential compared to [AuCl4]

– as does [AuBr 2]- in comparison

to [AuCl2]- . It can be deduced that reduction is harder for the Br complex than Cl

complex. In the growth solution, [AuBr 4]- exists as CTA-[AuBr 4]

- which is

expected to be even more stable. This is why a weak reducing agent such as

ascorbic acid (AA) cannot reduce the complex to gold atom while AA easily reduces

the [AuCl4] - to gold atom to produce spherical particles. AA can only reduce

[AuBr 4]- in the metallomicelles to [AuBr 2]

-. So the nucleation can be withheld until



26

the seed solution is added. With the catalytic action of the seeds, Au+1 ion in the

metallomicelle is reduced to gold atom. Therefore the effectiveness of CTAB comes

from the lower redox potential of Br complex.

To demonstrate the effect of Br substitution on the formation of NRs, KBr was

added to the growth solution prepared with CTAC. When KBr was added to CTAC,

short rods were obtained in significant fraction (Figure 2.4and Figure 2.5). This

result comfirms that the form of precursor is important in determining the morphology

of the particle.

400 600 800 10000.0

0.1

0.2

0.3


Wavelength(nm)

CTAC

CTAC+KBr

Figure 2.4. UV-Vis-NIR spectrum of gold NR solution. Effect of KBr on the

formation of NR in the growth solution containing CTAC.



27

Figure 2.5. A TEM image showing that significant amount of short rods are

obtained when KBr is added with CTAC. Scale bar is 50nm.

In synthesis of gold NPs, chloroaurate was dominantly used as starting gold

salt. From our results, it is found that chloride is substituted by bromide in the final

precursor form. We examined the possibility of using bromoaurate instead of

chloroaurate and indeed succeded to prepare gold NRs of aspect ratio equivalent to or

larger than those prepared by chloroaurate(Figure 2.6).



28

400 600 800 1000 1200

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07


Wavelength(nm)

HAuCl4

HAuBr 4

Figure 2.6. UV-Vis-NIR spectra of gold NR solution prepared by different gold salts.

Different morphology of gold NRs prepared by CTAB and CTAC may result

from different adsoprtion behavior of the two surfactant. Adsorption occurs initially

via attractive electorstatic interactions between the cationic headgroup and the anionic

surface of growing gold NPs. Once the original charge of the surface is neutralized

by the adsorption of oppositely charged surfactant ions, adsorption continues via

interactions between the hydrophobic tails, leading to surfactant self assembly on the

surface. Counterions affect the structure of this assembly by screening the

electrostatic repulsions between the ionic headgroups. It has been reported that

bromide ions have a 5-fold greater binding affinity for CTA+ than that of chloride

ions30,31. Above the CMC, the surface excess concentration of CTAB is 60% greater

than that of CTAC making the structure of the adsorbed surfactant as densely



29

packed wormlike micelle(Figure 2.7). Below the CMC, the counterion has only a

small effect on the srtucture of the adsorbed layer.

The structure of adsorbed CTAB on the surface of gold NRs was found to be

bilayer 32. The bilayer can result from the adsorption of the micelle on the surface as

suggested by the high resolution TEM study33. Considering the concentration of

CTAB in the growth solution is above CMC, we believe that the densely paked

wormlike micelle adsoption of CTAB on the growing gold NPs efficiently promote

the directional growth of particle resulting in the rod shape.

Figure 2.7. AFM images of adsorbed surfactant layers on silica observed at

10×CMC 30 . a) CTAB and b) CTAC.

The concentration of CTAB is the dominant factor to induce the anisotropic

shape of NRs. The range of concentration of CTAB used to prepare NRs is from

just above CMC to 0.1M in the literature. However, the effect of the concentration

of CTAB on the formation of gold NRs was not studied.

We found that as the concentration of CTAB increases, the yield of NRs

a) b)



30

increases which can be seen from the decreasing transverse surface plasmon (TSP)

intensity and increasing longitudinal surface plasmon (LSP) intensity (Figure 2.8).

d

c

b

a

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

400 500 600 700 800 900 1000

Wavelength(nm)


Figure 2.8. The effect of CTAB concentration on the formation of gold NRs. a:

[CTAB]=0.02M, b: 0.05M, c: 0.08M, and d: 0.1M.

This result can be explained by the enhanced CTAB adsorption on the surface

of the growing particle. The adsorption of surfactant onto an interface involves the

straightforward processes of diffusion of surfactant molecule (monomer) from the

bulk to the surface. The adsorption of monomer onto the surface disturbs the

monomer/micelle equilibrium and micelles can kinetically break up releasing

monomer which can then diffuse and kinetically adsorb onto the surface. Also,

micelles can directly adsorb onto the surface. Therefore, by increasing CTAB

concentration, the preferential adsorption on the growing gold particles can be



31

promoted to produce more NRs than spherical particles.

2.3.2. The role of BDAC in enhancing the formation of anisotropic shape of gold

NRs.

Figure 2.9. TEM images of gold NRs synthesized in a) CTAB and b) CTAB/BDAC.

Scale bar is 50nm.

Figure 2.9 shows the effect of BDAC for preparing longer NRs. When

BDAC is used along with CTAB, longer NRs were made. This result was attributed

to a more flexible template made by surfactant mixture and the different affinity of

CTAB and BDAC on the facet of growing gold NRs 20.

We took different approach to explain the role of BDAC by focusing the

precursor stage of the growth solution.

The precursor in the growth solution of CTAB and BDAC shows the

characteristic peaks of [AuBr 4]- instead of [AuCl4]

– , indicating that CTAB is used for

forming precursor (Figure 2.10).



32

0

0.5

1

1.5

2

230 330 430 530

wavelength(nm)

O p t i c a l D e n s i t y

CTAB+HAuCl4

BDAC+HAuCl4

CTAB+BDAC+HAuCl4


different surfactant. [HAuCl4] = 5×10-4 M for all solutions and [surfactant] = 5×10-2

M.

As previously mentioned, the solubilizability of the complexes can be mapped

at room temperature by visual inspection. To analyze the solubilizability

quantitatively, a series of vials containing HAuCl4 at constant concentration were

prepared and the concentration of surfactant was successively increased until the

turbidity disappeared. The solutions were sonicated for an hour after which the

UV-Vis spectra were taken. The change in turbidity can be monitored by the change

of absorbance. The absorbance at the wavelength 600 nm was chosen because this

peak is not related to any of the characteristic peaks of the solution. The absorbance

was plotted as a function of the concentration of surfactant. It is clearly shown in

Figure 2.11 that by using the binary system, the solution starts to solubilize the



33

complex at a much lower concentration than by using CTAB alone. This result is

consistent with the report that in a mixed surfactant system of BDAC and CTAB, the

first CMC was dropped from 0.001 to 0.0005M and the second CMC was also

dropped dramatically from 0.28 M to 0.0014M over all mixing ratios34.

0

0.02

0.04

0.06

0.08

0.1

0 0.02 0.04 0.06

Total conc.(mol/l)


CTAB

CTAB+BDAC

Figure 2.11. Change of the absorbance of the surfactant-HAuCl4 mixture solution at

600nm as a function of total surfactant concentration. The concentration of HAuCl4

was fixed at 5×10-4 M. The absorbance at 600nm is not related to any of the

characteristic peak of the solutions and is an indicative of the turbidity.

Since 2nd CMC is lowered by BDAC, the size of micelle is expected to become even

bigger through structural transitions from spherical micelle to rod shaped micelle35.



34

Therefore, we reasoned that the bigger micelles accomodate the precursor

facilely preventing immediate reduction which may lead to the formation of spherical

particles.

Since we found that the higher concentration is also effective in making longer

NRs in the case of binary surfactant system (Figure 2.12), we decided to investigate

the optimum ratio of CTAB/ BDAC at higher concentration. We fixed the BDAC

concentration to 0.125 M and narrowed down the range of ratio (between 0.2 and 1)

by changing the amount of CTAB.

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

350 550 750 950 1150 1350

Wavelength(nm)


d

a

b

c

Figure 2.12. UV-Vis-NIR spectra of gold NR solutions having different total

concentration of surfactant. The ratio of CTAB/ BDAC was 0.5. The total

concentration of surfactants was a) 0.135M, b) 0.180M, c) 0.225M, and d) 0.270M.



35

400 600 800 1000 12000.00

0.05

0.10

0.15

0.20


Wavelength(nm)

0.2

0.4

0.6

Figure 2.13. UV-Vis-NIR spectra of gold NR solutions with increasing ratio of

CTAB/BDAC when the concentration of BDAC is fixed at 0.125M.

Figure 2.14. TEM images of gold NRs with the ratio of CTAB/ BDAC a) 0.2, b) 0.4

c) 0.6, and d) 1. The concentration of BDAC is fixed at 0.125M. The scale bar is

100nm.



36

0

10

20

30

40

50

60

70

0 0.2 0.4 0.6 0.8 1 1.2

The ratio of CTAB/BDAC

T h e l e n g t h o f N R ( n m )

11

11.5

12

12.5

13

13.5

14

T h e d i a m e t e r o f N R ( n m )

length

diameter

Figure 2.15. The effect of the ratio of CTAB/BDAC on the size of NRs. The


increases, the length of NRs increases and the diameter of NRs decreases.

750

800

850

900

950

1000

1050

1100

0 0.5 1

the ratio of CTAB/BDAC

L S P ( n m )

Figure 2.16. The effect of the ratio of CTAB/BDAC on LSP of the growth solutions.

The concentration of BDAC is fixed at 0.125M. It can be seen that as the CTAB

fraction increases, the LSP of NRs reaches maximum at the ratio of 0.67 and

decreases. The reproducibility increases with increasing ratio.



37

From the UV-Vis-NIR spectra (Figure 2.13) and TEM images (Figure 2.14), it

can be seen that as the ratio was increased, the NRs became longer and uniform

(Figure 2.15), and the fraction of the byproduct significantly decreased. But on

further increasing beyond 0.6, the aspect ratio of NRs became smaller, as seen from a

shift of the LSP to lower wavelengths. Also reproducibility is increased by

increasing the amount of CTAB (Figure 2.16).

In this experiment, the concentration of BDAC was fixed and the

concentration of CTAB was changed to obtain given ratios. The amount of CTAB is

critical, because it is used as a precursor in making complex with gold ion and as a

stabilizer to confine and direct the growth of NRs. Therefore increasing aspect ratio

could be mainly due to the increasing concentration of CTAB.

b c

a

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

350 550 750 950 1150 1350

Wavelength(nm)


Figure 2.17. UV-Vis-NIR spectra of the gold NRs solutions with decreasing ratio of

BDAC/CTAB when the concentration of CTAB is fixed at 0.1M. The concentration

of BDAC is a) 0.03M b) 0.075M, and c) 0.125M.



38

To exclude the effect of the concentration of CTAB, a similar experiment was

done by fixing CTAB concentration and changing BDAC. Up to a certain point, as

the amount of BDAC increases, the aspect ratio increases (Figure 2.17). In both

cases, aspect ratio of NRs increases with increasing surfactant concentration and

shows a maximum at some surfactant concentration.

The number of micelles increases with increasing surfactant concentration.

When the number of micelles exceeds that required to solubilize all of the substrate,

there is a dilution of the concentration of substrate per micelle as the surfactant

concentration is further increased. This causes the drop in reduction rate. Also, the

number of monomer in the solution is increased by the increasing concentration of

surfactant which causes more efficient adsorption of surfactant on the growing surface.

Therefore more NRs can be made due to decreasing reduction rate and increasing

adsorption efficiency.

We conclude the role of surfactant in the formation of gold NRs as follows:

1) to make complex with gold salt which is more stable so that the reduction

renders slower.

2) to capture the complex inside the micelle so that it can protect gold ion

from being instantly reduced.



39

3) to adsorb on the surface of growing NRs so that it can confine the direction

of growth.

4) The role of BDAC as cosurfactant is to lower the 2nd CMC to promote

effective formation of metallomicelle and preferential adsorption on the

growing surface of gold NR.

2.3.3. The reduction

In the seed-mediated method, ascorbic acid (AA) is used as a mild reducing

agent. When a strong reducing agent (NaBH4) is used, the growth solution became

deep purple within several minutes. In this case, only nanospheres are produced

(Figure 2.17). When the stoichiometric ratio of AA was added without seed solution,

there was no color change for 24 hrs. But when the solution was kept for over a week,

pale purple color developed. But UV-Vis spectrum shows no significant NPs

formation. Only when AA was added to seed solution, NR was produced as a main

product.



40

400 600 800 1000 12000.00

0.02

0.04

0.06

0.08

0.10


Wavelength(nm)

NaBH4

Ascorbic acid

Figure 2.18. The effect of reducing agent on the formation of gold NRs.

From this result, it is apparent that weak reducing agent is essential for the

formation of rod shape. Also, the seed solution is a necessity as a growth reagent

inducing the second reduction process.

The reduction process of the gold ion by AA can be described as

First reduction : Au+3→Au+1

−+ +++−→+− Br 2H2OHCAuBr CTAOHCAuBr CTA 66626864

Second reduction : Au+1→Au0

−++ ++++→+− Br 4H2CTA2OHCAu2OHCAuBr CTA2 6666862

The first reduction is confined in the metallomicelles. The second reduction

only begins after the seed solution is added. The overall reaction is

−++

++++→+− Br 8H6CTA2OHC3Au2OHC3AuBr CTA2 6666864



41

We calculated the stoichiometry of this reduction reaction by considering the

ionization of AA and pH of the growth solution. That turns out to be 1.6. So for

the complete reduction, molar ratio has to be greater than 1.6.

Recently, the effect of AA concentration on the morpholgy of short NRs has

been reported36. The specific ratios of 1.6 and 7.5 were investigated. It was said

that when the ratio was high, the rod length decreased. But the reaction condition was

different for both of the experiments (1.6 with AgNO3 and 7.5 without AgNO3), thus

making it difficult to draw any general conclusions.

We studied the effect of the concentration of AA on the growth of NRs. In

particular, we investigated how the reduction rate affects the morphology of NRs.

The molar ratio of AA to gold ion was changed from 1 ( which is the ratio required to

reduce Au+3 to Au+1 ) to 1.9.

When the molar ratio was 1, no NPs were obtained. As the molar ratio

increased from 1.1 to 1.9, the absorbance at the peaks increased indicating higher

yield of particles formation. As the molar ratio increased from 1.1 to 1.3, LSP shifts

to longer wavelength but upon further increasing the molar ratio, LSP shifted back to

shorter wavelength (Figure 2.19). As it can be seen in the TEM images (Figure 2.20),

the initial red-shift can be related to the aparent increase in the length of NR resulting



42

in the increasing aspect ratio. Later the growth of the diameter overtook the growth

of the length resulting in decreasing aspect ratio.

When the ratio is above 1.6, the shape of NRs is no longer a spherocylinder.

The growth in diameter direction was greatest at the ends and progressively slower

towards center, giving a taper shape as shown in Figure 2.20(c).

0.00

0.05

0.10

0.15

0.20

0.25

350 550 750 950 1150 1350

wavelength(nm)


a

b

c

d

ef

Figure 2.19. UV-Vis-NIR spectra of 6 identical growth solutions in which the

ascorbic acid(AA) contents increased from a to f. The ratio of AA:gold ion was a) 1.1,

b)1.2 , c) 1.3, d)1.4, e) 1.5, and f) 1.6. The spectra were taken from the as-made

solutions after the growth was ended (measured in 1mm path quartz cell).



43

Figure 2.20.Changing morphology of NRs by varying the molar ratio of AA to Au+3.

a: 1.1, b: 1.3, and c: 1.6 (a′, b′, and c′ show the byproduct from the same solution).Scale bar is 50 nm.

Nanocubes were obtained as a byproduct. The fraction of nanocubes in the

solution increased from 17% to 50% as AA increased. The size of nanocube

increased continuously by increasing the amount of AA. The edge length reached

31nm. The broad transverse band as a result was caused by the size difference

between the width of NR and the edge length of nanocube. They contributed to the

plasmon bands at 512nm and 550 nm, respectively.

.



44

0 100 200 300 400 500 600

0.0

0.2

0.4

0.6

0.8

1.0

N o r m a l i z e d o p t i c

a l d e n s i t y

Time(min)

Molar ratio of AA/Au

1.1

1.3

1.5

1.7

1.9

Figure 2.21. The effect of ascorbic acid(AA) on the growth kinetics of gold NRs.

The molar ratio of AA to gold ion was changed from 1.1 to 1.9. The intensity at

LSP was taken and normalized by the final intensity.

0

10

20

30

40

50

60

70

80

1 1.2 1.4 1.6

the molar ratio of of ascorbic acid/HAuCl4

s i z e o f N P s ( n m )

length

end width

diameter

Figure 2.22. Effect of ascorbic acid(AA) on the size of NRs and nanocubes(NCs). The

lenght of the NRs first increases and then decreases with increasing amount of AA.

The diameter of NRs and the size of NCs increases as the amount of AA increases.



45

The reduction rate depends on the concentration of the reducing agent. As the

amount of AA increases, the reduction of gold ion becomes rather faster (Figure 2.21).

As a result, the reduction takes place before surfactants confine the growth direction.

Taper shaped NRs as well as the bigger nanocubes are obtained.

When the molar ratio of AA to gold ion is less than the stoichiometric ratio,

gold cations remain in the growth solution after the growth stops. The NR solution

was centrifuged. After taking the concentrated NR solution at the bottom, clear

supernatant was kept in a test tube. UV-Vis spectrum of this supernatant showed that

supernant contains very little amount of NRs. When AA was added to this

supernatant, the color of the solution changed indicating the formation of particles.

We expected that the reduction can be reinitiated by adding extra AA to the

NPs solution after the initial growth stops so that remaining gold ions can be further

reduced. We studied how the extra AA can reinitiate the reduction process and

change the size and shape of the original NPs.

The bulk NR solution was made. The ratio of AA to Au+3 was 1.1. After

48 hours, the solution was divided into 5 test tubes. Different amount of extra AA

was added to each test tube except one tube for comparison. As the amount of extra

AA increased, the optical density became higher. LSP shifted to longer wavelength

indicating the growth of NRs and TSP broadened as shown in Figure 2.23.



46

0.00

0.05

0.10

0.15

0.20

0.25

0.30

350 550 750 950 1150 1350

Wavelengths(nm)


b

a

c

d

e

Figure 2.23. UV-Vis-NIR spectra of the NR solutions before and after adding extra

ascorbic acid(AA) to the mother solution which was initially made with the molar

ratio of AA to Au+3 is 1.1 (a: mother solution, b:adding 10µl, c:30µl, d: 50µl, and e:

100µl).

The TEM images (Figure 2.24) show the growth of NRs in the length with

increasing AA. The width of NRs first increased a little and only the end of NRs

became fatter while the middle of NRs did not change much (Figure 2.24). The

broad transverse peak resulted from the bigger nanocubes and the fatter end of the

NRs. When the largest amount of AA was added, we observed bone shaped NPs

with split ends and boomerang shaped NPs. Nanocubes also became bigger and the 4

corners of nanocubes began to bulge (Figure 2.26). In all solutions, the fraction of

byproducts is 20% indicating little additional nucleation.



47

Figure 2.24. TEM images of gold NRs and the size distribution of each sample; a)

from batch solution, b) grown from the batch solution (10ml) by adding 30 ul of extra

AA, and c) grown from the batch solution (10ml) by adding 50 ul of extra AA. Scale

bar is 50nm.

0

10

20

3040

50

60

70

80

90

100

4 0 6 0 8 0 1 0 0

1 2 0

1 4 0

1 6 0

Dimension(nm)

F r

e q u e n c y

0

10

20

30

40

50

60

70

80

90

100

4 0 6 0 8 0 1 0 0

1 2 0

1 4 0

1 6 0

Dimension(nm)

F r e q u e n c y

0

10

20

30

40

50

60

70

80

90

100

4 0 6 0 8 0 1 0 0

1 2 0

1 4 0

1 6 0

Dimension(nm)

F r e q u e n c y



48

30

35

40

45

50

55

60

65

70

75

0 20 40 60 80 100 120

The amount of ascorbic acid added( )

L e n g t h o f N R ( n m )

5

7

9

11

13

15

17

19

D i a m e t e r o f N R ( n m )

µl

Figure 2.25. The growth of NRs by reinitiating the reducing process. The length of the

NRs was increased with increasing amount of extra AA added after the first growth

was completed. The width in the middle didn’t change significantly, while the width

in the end increases as the amount of AA increases.

Figure 2.26.TEM images of gold NRs grown from mother solution by adding extra

amont of AA. Bone shaped NPs as well as boomerang shaped NPs were obtained

when the largest amount of AA was added.

When the growth resumes by reinitiating the reduction by adding extra AA,

the length of NRs becomes longer and only the ends of NRs becomes bulged. This



49

happens because after the initial growth is completed, the side of NRs is mostly

adsorbed by surfactant which restricts the growth in diameter.

Therefore reduction mainly occurs at the end of NRs. When the largest

amount of extra AA is added, forced reduction makes the NRs take on the shape of

a bone37. We believe that facets consisting the end have different reactivity with

each facet having a distinct reduction rate different from other facets. Each facet

grows at a different rate so that we observe asymmetric branches with 2 or 3

branches.

It is interesting to compare the NRs made by adding the same amount of AA

but using different time scale. One solution is made by adding 100ul of 0.1M AA

initially while the other is made by adding 70ul of 0.1M AA in the beginning and

adding 30ul of 0.1M after 24 hrs. In the former case, the diameter of the NRs

gradually becomes bigger from the center to the end (Figure 2.27 a).

Figure 2.27. The different morphology of gold NRs. a) 100ul of 0.1M AA was

added initially, and b) 70ul of 0.1M AA was added in the beginning and 30ul was

added after 24 hrs. Scale bar is 50nm.

a) b)



50

The length of NRs is 65nm and the diameter in the middle is 20nm. In the

latter case, longer NRs with smaller diameter were obtained and only the ends of the

NRs became bulged (Figure 2.27 b).

Temperature is one of the major control parameters which affects the reduction

rate. To further investigate the effect of reduction rate on the morphology of gold

NRs, the reaction temperature was varied while all other experimental conditions

were kept the same. TEM images of gold NRs grown at different temperatures

between 20°C and 50 °C are shown in Figure 2.28. With decreasing temperature,

the aspect ratio increased from 3 to 6. It is worth noting that this increase in aspect

ratio is due to the decrease in diameter of the NRs as shown in Table 2. The

decreasing diameter of the NRs can be ascribed to the effective confinement of the

growth in diameter direction at low temperature which renders low reduction rate.

Figure 2.28. Changing morphology of NRs by varying the reaction temperature. a)

50°C, b) 30°C and c) 20°C. All other experimental parameters are the same. The

scale bar is 100nm.

a) b) c)



51

Table 2.2. Maximum LSP(λmax), length, diameter and aspect ratio of gold NRs

synthesized at different temperature.

Temperature(°C) λmax(nm) average

length(nm)

average

diameter(nm)

aspect ratio

20 732 46.8 7.9 6.030 954 53.4 10.0 5.6

50 1020 40.7 12.6 3.3

The reduction process in the synthesis of gold NRs can be summarized as

follows: Equimolar amount of AA is necessary to reduce Au+3 to Au+1 .

unconsumed AA reduces Au+1 to Au0 with the catalytic effect of the seed solution.

Hence, as AA concentration increases, NRs can be obtained in higher yield. The

molar ratio of gold ion to AA should be greater than 1.5 to ensure the complete

reduction of gold ion. The unreacted precursor can be reduced by reinitiating the

reduction which results in changing the morphology of existing NRs.

2.3.4. Tuning the aspect ratio of NRs by kinetically controlled growth

To obtain higher aspect ratio of NRs, we manipulate the reduction steps. First

we prepared mother solution using lower molar ratio (1.1). The aspect ratio is about 6

and the diameter is small. The solution contains a small amount of byproducts.

After the first growth leveled off we added extra AA every 150min to balance

the reduction and the adsorption. We repeated this 6 times (Table 2.3) and

successfully synthesized NRs with higher aspect ratios with mainly lengthwise growth

and little byproduct. Aspect ratio up to 9 can be obtained by this procedure.



52

Table 2.3. The modified reduction steps to obtain higher aspect ratio of gold NRs.

Addition

time(min)0 180 330 480 630 780 930

amount of AA(µl) 550 50 50 50 50 50 50

stoichiometricratio

1.1 1.2 1.3 1.4 1.5 1.6 1.7

LSP(nm)

(L/d)

1000

(6.4)

1040

(6.9)

1070

(7.3)

1080

(7.4)

1090

(7.5)

1120

(7.9)

1140

(8.2)

Figure 2.29. TEM images of (a) gold NRs from mother solution and (b) gold NRs

grown by adding small amount of ascorbic acid step by step. The scale bar is

100nm.

The main difference of controlling aspect ratio in our synthesis method is that

the procedure is rather simple. Usually obtaining higher aspect ratio of NRs often

involves the preparation of many growth solutions and growth can be done by taking

multiple steps which prevents to prepare low polydispersity NRs15. Separation of the

byproduct through many rounds of centrifugation is also required.

In our approach, the desired size and shape of the NRs was obtained in high

yield in a single growth solution by adding just more reducing agent step by step.

a) b)



53

2.3.5. The effectiveness of seed-mediated method

The advantage of seed-mediated method is to separate nucleation and growth.

This separation is supposed to minimize the further nucleation and decrease the size

distribution. We probed the effectiveness of seed-mediated method by preparing the

gold NRs without adding seed solution.

Instead of adding seed solution, we added a very small amount of NaBH4

solution to the growth solution (which is equivalent to the amount of NaBH4 in the

seed solution) expecting that a small amount of strong reducing agent will serve as

nucleating agent.

0.00

0.05

0.10

0.15

0.20

0.25

350 550 750 950 1150 1350

Wavelength(nm)


a

b

Figure 2.30. UV-Vis-NIR spectra of the NR solution grown without seed solution. A

very small amount of NaBH4 solution equivalent to the amount of NaBH4 in the seed

solution was added instead of adding the seed solution. In 24hrs, only short NRs were

produced and the yield was very low (a). After 1 month , they grew longer and the

yield became higher (b).



54

Figure 2.30 shows the UV-Vis-NIR spectrum of the growth solution over a

long time period. After 1 day, we observed that short NRs were formed with

significant amount of byproduct (50%). The NRs grew very slowly over 1 week and

became longer (aspect ratio 7).

Figure 2.31. TEM image of gold NRs grown without seed solution. The scale bar is

100nm.

The growth is slow because there should be an induction period to form nuclei

from which NRs can be grown. The amount of byproduct is higher than the seed

grown case because nucleation and growth take place at the same time. Figure 2.31

shows a TEM image of gold NRs grown by adding NaBH4 as a nucleating agent.

Since the growth is slow, the surfactant can effectively confine the growing direction,

the diameter of NRs is smaller than the diameter of NRs grown by adding seed

solution (average diameter of NRs grown without seed solution< 8 nm, average

diameter of NRs grown with seed solution > 8 nm). Therefore, the seed solution

must act as a catalyst for the reduction of AuBr 2- ions by the unconsumed AA due to

particle-mediated electron transfer from AA to AuBr 2- 38.



55

2.4. Conclusions

In summary, we have shown that the morphology of gold NRs is controlled

by manuplating the reduction rate and the adsorption kinetics of the surfactant. The

choice of surfactant is important not only because of its adsorption on the surface of

the growing particles but also because the precusor of gold-surfactant complex has

significant influence on determining ∆E of reduction reaction. Our results show

that the adsorption competes against the reduction of gold ion on the same surface of

the NR. When reduction rate is slower, long NRs with smaller diameter are

obtained while when reduction is forced, short NRs with larger diameter are obtained.

Therefore, by scheming the addition of AA in multiple steps, tuning the aspect ratio

and the shape of the NRs can be achieved.



56

References

1 Shalaev, V.M.; Cai,W.;Chettiar,U.K.; Yuan,H.; Sarychev,A.K.; Drachev,V.P.;

Kildishev, A.V. “Negative index of refraction in optical metamaterials” Optics

Letters 2005, 30, 3356.

2 a) Foss Jr., C.A.; Hornyak, G.L.; Stockert, J.A.;Martin, C.R. “Optical properties of

composite membranes containing arrays of nanocsopic gold cylinders” J. Phys.

Chem. 1992, 96 , 7497 b) van der Zande, B.M.I.; Bo1hmer,M.R.; Fokkink, L.G.J.;

Scho1nenberger,C. “Aqueous Gold Sols of Rod-Shaped Particles” J. Phys. Chem.

B 1997, 101, 852.

3

Pollard, A.P.; Heron, C. “Archaeological Chemistry”, Royal Society of Chemistry. 1996; Chapter 5.

4 Faraday, M.“Experimental Relations of Gold (and other metals) to Light” Philos.

Trans. 1857, 147 , 145.

5 Schmid, G.; Corain, B., “Nanoparticulated gold: Syntheses, structures, electronics,

and reactivities” European Journal of Inorganic Chemistry 2003, 17 , 3081.

6 Marie-Christine,D.; Didier,A.“Gold nanoparticles: assembly, supramolecular

chemistry, quantum-size-related properties, and applications toward biology,

catalysis, and nanotechnology” Chemical reviews 2004, 104, 293.

7 Turkevich, J.; Stevenson, P. C.; Hillier, J. “The nucleation and growth processes in

the synthesis of colloidal gold” Discussions of the Faraday Society 1951, 11 55.

8 Torigoe,K.; Esumi,K. “Preparation of Colloidal Gold by Photoreduction AuCl4--

Cationic Surfactant Complexes” Langmuir 1992,8 , 59.

9 Esumi,K.; Matsuhisa,K.; Torigoe,K. “Preparation of Rodlike Gold Particles by UV

Irradiation Using Cationic Micelles as a Template” Langmuir 1995,11, 3285.

10 Esumi,K.;Junko, H..;Nariaki, A.; Usui,K.; Torigoe,K. “Preparation of Anisotropic

Gold Particles Using a Gemini Surfactant Template” Journal of Colloid and

Interface Science 1998, 208 , 578.



57

11 Kameo,A.; Suzuki,A.; Torigoe,K.;Esumi,K. “Fiber-like Gold Particles Prepared in

Cationic Micelles by UV Irradiation: Effect of Alkyl Chain Length of Cationic

Surfactant on Particle Size” Journal of Colloid and Interface Science 2001, 241,

289.

12 Kim, F.; Song, J. H.;Yang, P. “Photochemical synthesis of gold nanorods” J. Am.

Chem. Soc., 2002,124, 14316.

13 Niidome,Y.;Nishioka,K.; Kawasaki, H.; Yamada, S. “Rapid synthesis of gold

nanorods by the combination of chemical reduction and photoirradiation processes;

morphological changes depending on the growing processes” Chem.Commun. 2003,

2376.

14 Yu,Y; Chang,S; Lee,C; and Wang, C. R. C. “Gold Nanorods: Electrochemical

Synthesis and Optical Properties” J. Phys. Chem. B, 1997, 101, 6661.

15 Jana, N. R.; Gearheart, L; Murphy,C. J.“Wet Chemical Synthesis of High Aspect

Ratio Cylindrical Gold Nanorods” J. Phys. Chem. B, 2001, 105, 4065.

16 Brown, K. R.; Walter, D. G.; Natan, M. “Seeding of Colloidal Au Nanoparticle

Solutions. 2. Improved Control of Particle Size and Shape” J. Chem. Mater . 2000,

12, 306.

17 Busbee, B.D.; Obare, S.O.; Murphy, C.J. “An improved synthesis of high-aspect-

ratio gold nanorods” Advanced Materials 2003, 15, 414.

18 Canizal, G.; Ascencio, J. A.; Gardea-Torresday, J.; Yacaman, M. J. “Multiple

twinned gold nanorods grown by bio-reduction techniques” Journal of Nanoparticle

Research 2001, 3, 475.

19 Mieszawska, A. J.; Zamborini, F. P. “Gold Nanorods Grown Directly on Surfaces

from Microscale Patterns of Gold Seeds” Chemistry of Materials 2005, 17,

3415.

20 Nikoobakht, B; El-Sayed, M. A. “Preparation and Growth Mechanism of Gold

Nanorods(NRs) Using Seed-Mediated Growth Method” Chem.Mater . 2003, 15,

1957.



58

21 Bakshi,S.; Kaur, I. “Head-group-induced structural micellar transitions in mixed

cationic surfactants with identical hydrophobic tails” Colloid Polym. Sci. 2003, 281,

10.

22 Jana, N. R.; Gearheart, L.; Murphy, C. J. “Seeding Growth for Size Control of 5-40

nm Diameter Gold Nanoparticles” Langmuir 2001, 17 , 6782.

23 Frens, G. “Controlled Nucleation for the Regulation of the Particle Size in

Monodisperse Gold Suspensions” Nature: Phys. Sci. 1973, 241, 20.

24 Goia, D.V.; Matijevic E. “Preparation of monodispersed metal particles” New

Journal of chemistry 1998, 22, 1203.

25 Zweifel,D.A.;Wei,A. “Sulfide-Arrested Growth of Gold Nanorods” Chem. Mater.

2005, 17, 4256.

26 Gao,J.; Bender,C.M.;Murphy,C.J. “Dependence of the Gold Nanorod Aspect Ratio

on the Nature of the Directing Surfactant in Aqueous Solution” Langmuir 2003, 19,

9065.

27 Veisz,B.; Kiraly,Z. “Size-Selective Synthesis of Cubooctahedral Palladium

Particles Mediated by Metallomicelles” Langmuir 2003, 19, 4817.

28 Puddephatt, R.J. “The chemistry of gold” Elsevier:Amsterdam-Oxford-New York,

1978

29 Bard,A. J. “Encyclopedia of Electrochemistry of the elements” vol. IV, Marcel

Dekker: New York and Basel, 1975

30 Velegol, S. B. ; Fleming, B. D.; Biggs,S.; Wanless,E. J.; Tilton, R. D. “Counterion

Effects on Hexadecyltrimethylammonium Surfactant Adsorption and Self-Assembly

on Silica” Langmuir 2000, 16 , 2548.

31 Atkin, R;. Craig, V.S.J.; Wanless, E.J. ; Biggs, S., “The influence of chain length

and electrolyte on the adsorption kinetics of cationic surfactants at the silica–

aqueous solution interface” Journal of Colloid and Interface Science 2003, 266 ,

236.



59

32 Nikoobakht, B; El-Sayed, M. A. “Evidence for bilayer assembly of cationic

surfactants on the surface of gold nanorods” Langmuir , 2001, 17 , 6368.

33 Gai, P.L.; Harmer, M.A. “Surface Atomic Defect Structures and Growth of Gold

Nanorods” Nano lett. 2002, 2, 771.

34 Bakshi, M. S.; Kaur,I. “Head-group-induced structural micellar transitions in

mixed cationic surfactants with identical hydrophobic tails” Colloid Polym. Sci.

2003 , 281, 10.

35 Ross, S.; Kwartler, C. E.; Bailey, J.H. “Colloidal association and biological activity

of some quaternary ammonium salts” Journal of Colloid Science 1953, 8 ,385.

36 Sau, T.K.;Murphy,C.J. “Seeded High Yield Synthesis of Short Au Nanorods in

Aqueous Solution” Langmuir 2004, 20, 6414.

37 Chen,S.; Wang,Z. L.; Ballato,J. ;Foulger, S. H.; Carroll,D. L. “Monopod, bipod,

tripod, and tetrapod gold nanocrystals” J. Am. Chem. Soc. 2003, 125, 16186.

38 Perez,J.; Gonzalez,E.R., “Hydrogen Evolution Reaction on Gold Single-Crystal

Electrodes in Acid Solutions” J.Phys.Chem.B 1998, 102,10931.



60

3. Chapter III

Morphology and Growth mechanism of Gold Nanorods

Abstract

Growth mechanism of gold NRs (NRs) was elucidated by studying the

crystallography of nanoparticles (NPs).

The seed particles grow in strong kinetics-controlled environment. They

are cubo-octahedra with a certain size distribution. Depending on the confinement

and the initial size of the seed particle, some of them grow to become NRs which are

enclosed mainly by 100 and 110 facets and their axial growth direction is [001]

and some grow to become truncated cubic having 100 and 111 facets.

Experimental images of gold NRs are shown that corroborate the proposed structure.



61

3.1. Introduction

In the wet chemical synthesis of NPs, the structures of particles are determined

both by the kinetic and thermodynamic factors. In most cases, a mixture of

different structures is obtained which represents the statistical thermodynamics of

the energies and the effects of kinetics during the growth process1. When there is

very strong kinetic control of the growth, it results in special types of particle

morphologies such as rods, needles, platelets, and dendrites.

This strong kinetic control is found in the synthesis of gold NRs where a weak

reducing agent and larger amount of surfactant are used for stabilizing unstable facet.

We reviewed the literature to summarize the key findings in the

crystallographies of gold NRs and how the specific structure is related to the growth

mechanism. The knowledge of the crystal structure of the particles may assist to

deduce the growing mechanism.

Characterization of gold NRs has been done mostly by the image of particle

and electron diffraction using transmission electron microscope (TEM).

Large fiber-like gold particles (2µm) obtained by UV irradiation at higher

concentration surfactant were found to have 110 faces while large gold particles

grown in the absence or low concentration of surfactant generally show 111 facets,

which have the lowest surface energy

2

. The preferential growth of the specific plane



62

is attributed to the adsorption of the surfactant molecules.

Wang et al. studied gold NRs synthesized electrochemically using surfactant

micelles as a capping material using high resolution transmission electron microscope

(HRTEM)3 and found that short gold NRs with aspect ratios of 3-7 were single

crystals without dislocation, stacking fault or twin enclosed mainly by 100 and

110 facets with a [001] axial growth direction(Figure 3.1).

Figure 3.1. The structural model of gold NRs proposed by Wang et al. 3

They also found that longer gold NRs with aspect ratios 20-35, which are

observed occasionally, contained single twinning. Their surfaces were

predominantly 111 and 110, and their axial growth direction was <112>.

Spherical particles with mass equivalent to the short rods, was found to be truncated

octahedra, icosahedra, or decahedra with 111 and 100 facets and multiple

twinning.

Kim et al obtained gold NRs by UV irradiation and found that the



63

crystallographic facets are the same as the electrochemically synthesized gold NRs,

with the growth direction being [001] and the side mostly covered with 001 and

110 facets4.

A seed-mediated sequential growth process has been taken by many research

groups as a synthesis method to obtain high aspect ratio of NRs. In the observation

of gold NRs prepared by this method, Johnson et al.5 suggested that the gold NRs had

a penta-twinned structure with five 111 twin boundaries arranged radially to the

<110> direction of elongation and ten 111 end faces. The side faces were

believed to be either 100 or 110, or both. Gai et al6 also suggested that NRs are

multiply twinned particles (MTPs) with (110), (111) faces on sides and [100]

direction.

Figure 3.2. The structural model of gold NR proposed by Gai et al.6

One can notice that some of the fcc metal NRs have been claimed to be single

crystals, while others are composed of multiple crystal units. The apparent

discrepancy may originate from the different synthesis methods which result in

different kinetics of growth. For example, in seed-mediated sequential method, NRs

are synthesized by multiple steps such that usually seed solution is added to the



64

growth solution and this growth solution becomes seed solution and again added to

the new growth solution. It takes several hours to complete the whole process of

growth. The final size of NRs which is used in the characterization of crystallography

is usually in the range of hundreds nanometers in length. By contrast,

electrochemically grown NRs can be obtained in an hour and the size is tens of

nanometer.

The presence of the 110 facets is the common observation in the literature.

It is a unique feature of gold NRs and is not usually observed in spherical NPs

because the 110 facet has higher energy than that of the 111 and 100 facets.

It is necessary to examine the mechanism and link the mechanism to the

resulting crystallography of NRs. Several possible growth mechanisms have been

recently proposed 6-8.

Zipping mechanism7 was proposed focusing on the role of CTAB. Gold

atoms are added to the end faces while gold-surfactant complexes are added to the

side of the rods where CTAB binds as a bilayer to the growing sites assisting NP

elongation. This mechanism assumes the pre-existence of rod shaped particles and

fails to explain the formation of significant amount of nanosphere.

Gai et al. proposed that multiply twinned particles are made prior to the rod

formation, which have twin boundaries

6

. Weaker bonding in twin boundary areas



65

can provide sites for agglomeration. Leontides et al.8 reported the synthesis of

thread-like gold particles by UV irradiation as the reduction path and proposed that

the gold-surfactant pairs are solubilized in the micelles and intially formed particles

aggregates directionally by preferential surfactant binding.

These proposed mechanisms have difficulties explaining several aspects of the

observations made in the synthesis. For example, they fail to explain the diversity of

shapes produced in a single synthesis reaction.

Therefore,we need to approach the mechanism of gold NRs formation not by

considering only rod shape but considering all the possibility to form the different

structure and to elucidate why the formation of rod shape is preferred.

We prepare gold NRs by the one step seed-mediated method in binary

surfactant system and attempt to elucidate the growth mechanmis by examining the

crystallographies of gold NRs and sphercal particles.

3.2. The crystallographic model of gold nanorods

Before proceeding to discuss crystallography of gold nanorods, we want to

point out some issues in TEM observation.

TEM is the classical way to determine the size and shape of the NPs. But

there are a number of drawbacks to the TEM technique. Many materials require



66

extensive sample preparation to produce a sample thin enough to be electron

transparent, and changes in the structure may be caused during this process. Also

the field of view is relatively small, raising the possibility that the region analyzed

may not be characteristic of the whole sample, which leads to bad statistics9. Also, it

has the limitation such that it only provides a projected image, and although there is

some depth data, this is not cleanly interpretable.

We demonstrated how data can be biased depending on the field of view.

Figure 3.3 shows 3 images taken from the same TEM grid in different field of view.

The size of NRs was measured from each image was tabulated in Table 3.1 showing a

significant differences depending on the field of view and the sample size.

Figure 3.3. The TEM images of gold NRs taken from the different field of view of a

specimen, showing different aspect ratios.

It is known that NPs are self assembled especially driven by evaporation of the

water. This process is sure to occur in the process of preparing the TEM grid.

Therefore, measuring some (200) particles from subjectively selected spots and

determining the size of the NRs cannot be the legitimate method. TEM observation

has to be accompanied by other methods such as UV-Vis-NIR absorption



67

spectroscopy, light scattering and size exclusion chromatography

Table 3.1. Particle size analysis of gold NRs using different sampling area and number

of sample.

Figure 3.4 illustrates the collection of different shapes and sizes of NPs

produced by the synthesis method we used. For NRs, the shape is spherocylinder

for all aspect ratios. The ends of the NRs don’t have sharp edges. It rather has a flat

end with tapered side.

Figure 3.5 shows the electron diffraction from a single rod. From the

diffraction contrast, it is clear that gold NRs have the single crystalline structure

regardless of the aspect ratio. Electron diffraction pattern supports the facets enclosing

the NRs are 110 and the growth direction is [001].

Figure 3.4. The collection of different shape and size of NPs produced by the one step

seed mediated synthesis in binary surfactant system. a) aspect ratio of 3, b) aspect

ratio of 5, and c) aspect ratio of 7.

count (N) length(nm) diameter(nm) aspect ratio

A 100 70.6±12.1 15.4±1.5 4.5±1.0

B 100 85.4±20.1 15.5±1.9 5.5±1.9

C 100 81.5±11.6 15.5±1.3 5.3±1.0

D 300 80.2±15.2 15.6±1.6 5.2±1.3



68

Figure 3.5. The electron diffraction pattern of gold NRs from a single rod. a) L/d= 24

and b) L/d= 4. Scale bar is 50nm.

For isomeric NPs, polygonal shape of particles and cubic particles are found

as byproduct depending on the synthesis condition (the ratio of CTAB/BDAC, the

concentration of surfactant). The big spherical particles (>50nm) exist in a very

small fraction and appear to be MTP (Figure 3.6 a). Figure 3.6 b shows the

truncated cubic particle.

Figure 3.6. TEM image of NRs along with bigger MTP and cubic as byproducts.

Since the majority of NPs we prepared are single crystal, the models based on

MTP are not relevant.

a) b)



69

TEM observation has the limitation of providing only the projected image of

particle. However, we were able to obtain the information of cross section through

TEM. We prepared TEM grid with highly concentrated NRs solution to study self

assembly of NRs. Upon assembly of the rods into a nematic phase, we found areas of

what appears to be defects that are characteristics of nematic fluids. Upon looking at

the core of these defects we found what appear to be spheres, however, on tilting it is

clear that they are not spheres, but rather rods standing up on their ends.

Thus we were able to visualize the cross section of NRs. We found that the

cross section is octagon shape. Figure 3.7 shows octagonal cross section of the NRs

packed as a square packing.

Figure 3.7. Octagonal cross section of the NRs packed as a square packing.

The TEM observation of NRs is consistent with Wang’s model3 where

single crystal gold NR is enclosed mainly by 100 and 110 facets and its axial

growth direction is [001].



70

3.3. Growth mechanism

Then the question is which growth mechanism leads to this morphology.

Since we choose the seed-mediated synthesis method, the crystalline structure of the

seeds at the nucleating stage is critical for guiding the shapes of NPs due to its

characteristic unit cell structure. Also, it is important to know the interface between

the seed and growing materials, which has not been well resolved yet.

There have been several reports that the size of seed made by NaBH 4

reduction is about 4nm10 , but neither direct TEM observation nor the information

about the crystal structure of seed particle is available since the size is too small.

Though, it is reported earlier that small single crystal of gold is of roughly wulff

construction shape with (111) and (100) facets 1.

Growth rates are controlled by the sticking probability on a given face. Since,

for instance a (001) face of an fcc material has a higher sticking probability, being of

higher surface free energy than a (111) face, it will grow faster. The kinetic structure

will therefore be dominated by slower growing faces such as (111), and might contain

no (001) facets.

When the shape of the seed is assumed as a cubo-octahedral particle which has

8 hexagons of (111) surface and 6 square shapes of (100) surfaces, we can expect the

same results. Since (111) surface is more stable than (100) surface, growth will take



71

place mainly in <100> direction. When the growth rate is the same for every <100>

direction, then the particle will become truncated cubic which we observed in our

synthesis.

Figure 3.8 a) Possible growth mechanism of gold nanocube. b) Resulting gold cubic

particles. Scale bar is 50nm.

However, preferential adsorption of molecules onto one or more of the surface

planes can significantly alter the relative stabilities of different planes. Assuming

the case when the growth rate is low and the growth rate of each direction is not all

the same. While growing to <100>, there is chance to have (110). If this

metastable (110) face is stabilized by the surfactant, it will promote the faster growing

in that specific direction of <100>, hence rod shaped particle formation can be

envisioned.

The relation between the diameter and length of NRs may provide another key

to approach the growth mechanism. The general trend in our results shows that

longer NRs have smaller diameter. Figure 3.9 shows the different aspect ratios of

NRs synthesized in one batch. It is clearly shown that longer NRs have smaller



72

diameter.

Figure 3.9. TEM image of NRs synthesized in one batch. Scale bar is 50nm.

Figure 3.10. a) The relation between diameter and length of NRs. b) The relation

between aspect ratio and volume of NRs.

The calculation of volume shows that there is no correlation between the

aspect ratio and the volume of NRs. We reasoned that initially smaller seed becomes

longer NR and initially larger seed becomes shorter NR in a given period of time. The

amount of gold reduced on a given particle is almost the same.

We synthesized the NRs without seed solution. Instead of adding seed

solution, a very small amount of NaBH4 solution equivalent to the amount of NaBH4

in the seed solution was added. Figure 3.11 shows the TEM image of NRs we

obtained from this synthesis. Comparing the dimension of NRs, we found that the

a) b)

8

9

10

11

12

13

14

30 40 50 60 70 80

length (nm)

d i a m e t e r ( n m )

.

3000

4000

5000

6000

7000

8000

0 1 2 3 4 5 6 7 8 9

aspect ratio

v o l u m e ( n m 3 )

.



73

diameter of NRs obtained from this synthesis is smaller than that of the seeded

method for all aspect ratio.

Figure 3.12 shows the relation between the length of NRs and the diameter of

NRs. The longer the NRs, usually the thinner the NRs. The NRs from seedless

synthesis shows the same trend but the diameter is smaller. For the seedless

synthesis, nucleation and growth cannot be separated. As the primary particle forms,

it grows to NRs . Therefore the preferential adsorption can be involved from the

initial stage of growth to enduce the thinner NRs.

Figure 3.11. TEM image of gold NRs grown without seed solution. Scale bar is 100

nm.



74

5

6

7

8

9

10

11

12

13

14

30 50 70 90 110

Length (nm)

D i a m e t e r ( n m

)

Figure 3.12. The relation between length and diameter of NRs. Open circle

represents the NRs prepared without seed and filled square represents the NRs

prepared with seed. The diameter of NRs without seed is smaller when the length is

similar.

We propose the following as the growth mechanism for gold NRs.

Step I. Formation of gold-surfactant complex

CTA-Br + HAuCl4→CTA-[AuClmBr 4-m] (m=0-4)→ CTA -[AuBr 4]

Au+3 is reduced to Au+1 by ascorbic acid resulting in CTA -[AuBr 2].

Step II. Solubilization of the complex by micelle

Metallomicelles are formed. gold-surfactant complex is protected

from spontaneous consumption.

Step III. Formation of primitive NPs

The primitive cubo-octahedral NPs are produced. . CTAB molecules

start to be adsorbed on the surface of the particles.

Step IV-1. Growth of NRs and nanocubes

Gold-surfactant complex is transferred from the micelle to the growing

site. Gold ion diffused and is reduced by ascorbic acid with the



75

catalytic effect of seed. At the same time, CTAB molecule is

adsorbed on the surface of NP joining the existing layer of CTAB due

to hydrophobic interaction between the long carbon chains. The layer

of CTAB stabilizes the elongated NPs. As CTAB forms the first

layer on the surface of NPs, free molecule of CTAB in the solution

starts to be adsorbed to the existing layer of CTAB forming bilayer.

Depending on the efficiency of stabilization of metastable face of (110),

the faster growing in that specific direction of <100> results in the

formation of NRs. When the growth rate is the same for every <100>

then particle will become truncated cubic which we observed in our

synthesis.

Step IV-2. Growth of spherical shape NPs

Some of the primitive particles aggregate to become spherical shape of

NPs. Since the aggregation is random, the size is not uniform.

What can be explained by this growth mechanism?

• The effect of concentration of surfactants: when the concentration is lower,

the size and number of micelle are not enough to solubilize Au-surfactant

complex to protect them from spontaneous reduction. As the concentration

increases, more complexes are protected and released to the growing site over

longer period of time.

• The existence of optimum ratio of mixing BDAC and CTAB: when BDAC

and CTAB are mixed at the optimum ratio, the micelle size is maximized for

protecting more complex inside the micelle. And the number of free CTAB

molecule in the solution is also maximized to facilitate absorption of

surfactant on the surface of NPs.



76

• The formation of NRs as well as byproducts: the shape and size of the NPs are

decided by the size of primitive particle to begin with.

• The formation of nanobone: after the growth stops, CTAB molecule

continuously is absorbed on the surface of NP to make dense bilayer. When

the growth resumes, the reduction is reinitiated only at the ends of NPs

resulting branched shape.

3.4. Conclusions

In summary, we describe the growth mechanism of gold NRs elucidated by

crystallography of NPs. The particle shape of the seed corresponds to cubo-

octahedron. The seeds which grow in strong kinetics controlled environment

already have a certain size distribution. Depending on the confinement and the

initial size of seed particle, some of them grow to become NRs which are enclosed

mainly by 100 and 110 facets and their axial growth direction is [100] and some

grow to become truncated cubics having 100 and 111 facets.



77

References

1 Marks, L. D. “Experimental studies of small particle structures” Rep. Prog. Phys.

1994, 57 , 603.

2 Kameo, A.; Suzuki,A.;Torigoe,K.; Esumi,K. “Fiber-like Gold Particles Prepared in

Cationic Micelles by UV Irradiation: Effect of Alkyl Chain Length of Cationic

Surfactant on Particle Size” Journal of Colloid and Interface Science 2001, 241,

289.

3 Wang, Z.L.; Mohamed, M.B.; Link, S.;El-Sayed, M.A. “Crystallographic facets and

shapes of gold nanorods of different aspect ratios” Surface Science 1999,440,L809.

4 Kim, F.; Song, J. H.;Yang, P.”Photochemical synthesis of gold nanorods” J. Am.

Chem. Soc. 2002,124, 14316.

5 Johnson, C.J.; Dujardin, E.; Davis, S.A. ; Murphy, C.J.; Mann, S. “Growth and

form of gold nanorods prepared by seed-mediated, surfactant-directed synthesis”

Journal of Materials Chemistry 2002, 12, 1765.

6 Gai, P.L.; Harmer, M.A. “Surface Atomic Defect Structures and Growth of Gold

Nanorods” Nano lett. 2002, 2, 771.

7 Gao,J.; Bender,C.M.;Murphy,C.J. “Dependence of the Gold Nanorod Aspect Ratio

on the Nature of the Directing Surfactant in Aqueous Solution” Langmuir 2003, 19,

9065.

8 Leontidis,E.;Kleitou,K.;Kyprianidou-Leodidou, T.; Bekiari,V.; Lianos, P. “Gold

Colloids from Cationic Surfactant Solutions. 1.Mechanisms That Control Particle

Morphology” Langmuir 2002, 18 , 3659.

9 Schnabel,U; Fischer,C.; Kenndler,E.“Characterization of Colloidal Gold

Nanoparticles According to Size by Capillary Zone Electrophoresis” J.

Microcolumn Separations, 1997, 9, 529.


Nanorods(NRs) Using Seed-Mediated Growth Method” Chem..Mater . 2003, 15,

1957.



78

4. Chapter IV

Shape and Size Separation of Gold Nanorods

Abstract

The shape separation of gold nanorods from a mixture of rods and spheres is

accomplished by centrifugation. We elucidate the hydrodynamic behavior of

nanoparticles of various shapes to identify the parameters to achieve efficient

separation. We describe the relative sedimentation behavior of colloidal rods and

spheres by the ratio of the sedimentation coefficients between rods and spheres where

the relative diameter of the particles determines the relative sedimentation of rods and

spheres for a given aspect ratio.



79

4.1. Introduction

Nanoparticle synthesis is characterized by a polydispersity in shape and size.

The physical and chemical properties as well as the applications of nanoparticles are

controlled and limited by their dimensions and shape1 .

As significant research efforts have been directed to optimize the synthesis

conditions2, there have been efforts to separate nanoparticles of different sizes and

shapes, since most of the synthesis methods result in mixture of particles with

different morphologies.

Size selective precipitation has been commonly used for nanospheres

dispersed in organic solvents 3 , 4 . Capillary electrophoresis 5 and column

chromatography6 have been used in the separation of metal nanoparticles.

In gold nanorod(NR) synthesis, the production of spherical particles as

byproduct is inevitable. Depending on the different methods, the fraction of

byproduct can be 10% to almost 90 %7,8,9. The use of separation methods mentioned

above is limited because some methods are not applicable and some methods only

result in partial separation10.

Regular centrifugation has been a part of the routine procedure to concentrate

nanoparticle solution and to separate excess capping material from the solution.

Also under certain conditions, separation of particles with significant mass difference



80

has been achieved 8 . The advantages of the separation by centrifugation are ease of

procedure and applicability to large-scale samples.

One can find the centrifugation parameters such as acceleration and duration

mentioned in the scattered reports7,8,11. The centrifugation methods can be classified

into 2 cases. When the size of NRs is less than 100nm in length, as-made solution is

centrifuged at lower rpm (2000 to 6500rpm). Then, the supernatant containing NRs is

centrifuged at much higher rpm (>10000rpm) to obtain precipitate at the bottom (case

1)7,11. When the size of NRs is much bigger, lower rpm(2000rpm) is used for getting

these big particles at the bottom (case 2)8. Through the repeated procedure,

relatively monodisperse gold NRs can be separated from small size spheres and NRs.

However, we failed to achieve efficient separation repeatedly and reproducibly

using those parameters in the literature7,8,11 because, depending upon cases, small

differences in concentrations, particle dimensions and centrifugation parameters drove

rods either to sediment out or remain in solution.

Centrifugation in a sucrose gradient, having been used in biological

application, was adopted to separate spherical particles from rod-shaped particles12.

Separation was conducted in a solution of sucrose where the density of the solution

increases from the top to the bottom of the centrifuge tube. The particles at the

bottom part of the tube are spherical particles and NRs were sorted based on their



81

mass with increasing aspect ratio from the top to the bottom tube. Faster

sedimentation of spherical particles and slower sedimentation of NRs were attributed

to the volume effect which affected the buoyancy. Although the experiment was

done systematically, the theoretical explanation was not entirely satisfactory.

Jana observed precipitation of NRs from a mixture of rods, spheres and plates

with different sizes13. This observation was made only when the solution of the

mixture was concentrated and the extra CTAB was added. The precipitation of NRs

on the bottom of the flask was attributed to the surfactant assisted self assembly of

NRs without further theoretical considerations.

Driven by the need and desire to understand these separation phenomena, we

set out to identify the physics that drives shape separation in centrifugation. We aim to

investigate the role of hydrodynamics in the centrifugation of Brownian rods and

spheres, to determine the size, shape and concentration dependence of sedimentation

coefficients and velocities, and to establish conditions that optimize shape separation.

The analysis was performed in collaboration with Vivek Sharma, and successfully

interprets and explains the experimental observations of centrifugation.



82

4.2. Theory

We consider a simple phenomenological analysis of sedimentation-diffusion

equilibrium14. The forces exerted on the particles during centrifugation include

centrifugal force, rmF 2

c ω = , buoyant force, o

2

b rmF ω −= , Brownian fluctuating force,

F f , and viscous drag force, fvF d −= , where ω is the centrifugation speed in rpm, m is

the mass of particle, mo is the mass displaced by the particle, r is the distance from the

center to the location of particle, f is the friction or drag coefficient, and v is the

sedimentation velocity. The balance of these forces leads to the Langevin equation of

particle undergoing Brownian motion under the influence of an external force. In the

stationary state, 0F F F F F f cbd total =+++= yields the Svedberg

coefficient, f / )mm(r / vS o2 −== ω , which expresses sedimentation velocity per

unit of applied centrifugal acceleration. Svedberg coefficient is a measure of

sedimentation rate and depends upon the ratio of effective mass and friction factor.

Next, we need to compute f, which is shape dependent, and compare the

sedimentation coefficients of rods to those for spheres. The relative importance of

thermal diffusion and flow is judged by a dimensionless group called Peclet number

(Pe), which is the ratio of the time a particle takes to diffuse a distance equal to its

diameter D to the time it takes to sediment this distance. Pe for typical Brownian

nanosphere is 1/)(/

2

<−== T k ramm DvaPe Bo ω , where a is the radius of the



83

sphere, and D is the thermal diffusivity, implying that the thermal fluctuations are

non-negligible. This distinguishes nanoparticle sedimentation theories from the

theories applied to macroscopic falling objects while the low Reynolds number

implies that the inertial effects are negligible and the Stokes or creeping flow

equations apply15

.

Drag on a Stokesian sphere is 6 πη av, where η is the viscosity of the solution,

and a is the radius of the sphere, implying a friction coefficient, f=6 πη a

and η ρ ρ 9 / a)(2S 2

o

sph

0 −= , where ρ is the density of the sphere and ρ0 is the

density of the solvent. In fact, the friction coefficients are similar for spherically

isotropic objects, hence the description of dynamics of spheres holds equally well for

the cubic, icosahedra and octahedral bodies encountered during synthesis of colloidal

particles16

. For anisotropic bodies, both force and torque balances are required, and

coupling of translation and rotation needs to be considered. Translational friction

coefficient of a falling single rod depends on the orientation of rod, and the friction

felt parallel to the rod is ½ that of the transverse falling rods, i.e //2ζ ζ =⊥ , and

[ ] oo

rod

o d Ld S η υ υ ρ ρ 24/)()/ln(2)( //

2 +−−= ⊥ where L is the length of the rod, d is the

diameter of the rod, and ⊥υ and // υ are the correction factors of the rod

perpendicular or parallel to the rods orientation, respectively. The actual values of

⊥υ and // υ for cylindrical rods are equal to -0.84 and 0.21, respectively

17

.



84

Thus in calculating drag for the rods, the orientation of the rods plays as big a

role as the dimensions of the rod itself.

Relative sedimentation behavior of colloidal rods and spheres can be described

by the ratio of the sedimentation coefficients between rods and spheres.

)](d / Lln2[)a2 / d (6 v / vs / s //

2

o

sph

o

rod

o

sph

o

rod

o υ υ β +−=== ⊥

This ratio for single rod and single sphere allows us to see that, for this case,

the central role in separation is played by the square of the ratio of the diameters of

the rod and sphere. For a given L/d, whether the rods or spheres sediment faster is

mainly controlled by the relative diameters of the particles, and since the aspect ratio

dependence enters through the log term, the effect is much less dramatic than the

effect of the diameters.

We plot the ratio of the sedimentation coefficient between the rod and the

sphere (β0) as a function of the ratio of the diameter of rod and sphere for different

aspect ratios of the rods (Figure 4.1). We fix the diameter of sphere as 20nm and

increase the diameter of rod from 5 nm to 20 nm to obtain the ratio ranging from 0.5

to 1. Depending on the predetermined aspect ratio of rod, the length of rod is

calculated by multiplying aspect ratio and the diameter of rod.

When the diameter difference between sphere and rod is small, β0 largely

depends on the aspect ratio and it is obvious that NRs with higher aspect ratio



85

sediment faster than spheres. In case that the diameter of sphere is larger than that of

NR, β0 becomes much smaller than the previous case indicating higher possibility of

getting faster sedimentation of spheres.

The experimental parameters shown in the literature could be represented by

the marked regimes in the plot. When lower rpm was used to separate NRs by

taking supernatant as a result of the faster sedimentation of nanospheres, the aspect

ratio of NRs is less than 5 and the ratio between the diameter of rod and sphere is less

than 17,9,11 . This condition can be mapped in regime I.

When the aspect ratio of NRs is much larger (>15) and the diameter of NRs is

comparable with that of spherical particles, lower rpm is used for getting these big

particles at the bottom8 because centrifugation results in the faster sedimentation of

NRs than spherical particles. This condition can be mapped in regime II.



86

L/d=20 6

5

0

5

10

15

20

25

30

35

40

45

0 0.2 0.4 0.6 0.8 1

d/2a

β 0

I

II

Figure 4.1. The ratio of the sedimentation coefficients between rod and sphere as a

function of the ratio of diameters between them. The higher the ratio is, the higher

the chance for rod sediment faster.

Hence, the presented analysis successfully interprets and explains the experimental

observations of centrifugation.

In this chapter, we validate this analysis by our own results. We further

investigate the parameters affecting the efficient separation by centrifugation.

4.3. Experimental

Gold NRs with different aspect ratios were synthesized according to

procedures described in chapter 2.

4.3.1. Centrifugation of as-made solution

As-made gold NR solution was centrifuged at 5600 G for 20 to 40 min (Jouan



87

centrifuge MR23i). After the centrifugation, the tube was visibly inspected.

The concentrated dark solution at the bottom (inside) of the tube was taken

using syringe. The supernatant was carefully removed. Solid-like deposit on the side

wall was dissolved by adding distilled water.

The yield of obtaining NRs was calculated based on the intensity of the LSP

and the volume of NR solution obtained.

mm

p p

V AV Aseparationtheof Yield

××=

where Am is the absorbance at longitudinal surface plasmon peak of the mother

solution, A p is the absorbance at longitudinal surface plasmon peak of the purified

solution, V m is the volume of the mother solution and V p is the volume of the

purified solution.

4.3.2. Precipitation of NRs in gravitational field.

As-made NRs solution (or the solution of interest) was kept in an Erlenmeyer

flask at 20°C. After 24hrs, the supernatant solution was taken and distilled water was

added to the flask to dissolve the sedimentation.

UV-vis spectra were acquired with a Cary 5G UV-visible-near IR

spectrophotometer. The morphology and the mean size of NPs were examined by

TEM (JEOL100 at 100 KV). For each sample, the size of more than 500 particles in

the TEM images were measured to obtain the average size and the size distribution.



88


4.4.1. Separation of nanorods from spherical particles.

Figure 4.2 shows the TEM images taken from the mother solution showing the

mixture of rods and spherical particles. The average diameter of the spherical particles

is 16.59 nm while the average diameter of NRs is 8.06nm. The average length of

NR is 58.54 nm. Hence, the aspect ratio is 7.3. The number fraction of spherical

particles is less than 10% (counting 700 particles) which is confirmed by the UV-Vis-

NIR spectroscopy where the intensity of transverse surface plasmon peak (TSP) is

low (Figure 4.3).

.

Figure 4.2. TEM image of the nanoparticles in the mother(as-made) solution, showing

coexistence of the nanorods and nanospheres. L/d of gold NRs is 7.3.



89

400 600 800 1000 1200

0.00

0.02

0.04

0.06

0.08

0.10

0.12


Wavelength(nm)

Figure 4.3. UV-Vis –NIR spectrum of the mother solution. L/d of the gold NRs is

7.3.

Figure 4.4 schematically shows the centrifuge tube after the centrifugation. A

fixed angle rotor was used in this experiment. In a fixed angle rotor, since the tube is

inclined, the materials are forced against the side of the centrifuge tube, and then slide

down the wall of the tube. Therefore deposition on the bottom is what settles down

faster. The deposition on the wall is what is later settled down. The color of the

solutions taken from the bottom and the side wall of the centrifuge tube is

distinctively different, which indicates that the particles deposited in different sites

have different optical property due to the different size and shape.



90

Figure 4.4. a) Schematic drawing of a centrifuge tube after the centrifugation and the

color of resulting solutions. b) the color of the solution taken from two different

locations shown in (a).

The UV-Vis-NIR spectrum (Figure 4.5) of the solution of the NPs from the

side wall shows intense longitudinal surface plasmon peak (LSP), red shifted from the

mother solution, and very weak TSP, indicating that it contains mostly NRs and the

aspect ratio is larger than that of the mother solution. The spectrum of the solution

of the bottom shows broad TSP with high intensity indicating that it contains mostly

spherical particles and some NRs. The LSP of this solution is blue shifted from the

mother solution, indicating that the aspect ratio of NRs is smaller than that of the

mother solution.

a) b)

Bottom Side



91

400 600 800 1000 12000.00

0.05

0.10

0.15

0.20

0.25

0.30


Wavelength(nm)

side wall bottom

Figure 4.5. UV–Vis-NIR spectra of the solutions of the deposits on the side wall of

the tube and at the bottom.

The size of the particles in these two solutions was measured by TEM

(counting 500 particles from each case). TEM images (Figure 4.6) show that indeed

the separation of NRs was successfully done by centrifugation. The bottom solution

contains lots of spherical particles, while the solution from the side wall contains

mostly NRs. In the NRs solution we prepared , the diameter of spherical particle is

almost twice larger than that of NRs, resulting the smaller ratio of sedimentation

between rods and sphere (average diameter of sphere is 16.6 nm). Therefore, the

spherical particles are sedimented faster than the rods. We compared the dimension

of NRs before separation and after separation (Table 4.1). Separated NRs from the

side wall have longer length and smaller diameter thus higher aspect ratio than those

in the mother solution. Shorter NRs were found in the bottom solution.



92

Figure 4.6. TEM images of gold nanoparticles: a) mother solution, b) after

centrifugation, NRs deposited on the side wall of the tube, and c) after

centrifugation, sedimented at the bottom, nanocubes, spheres and NRs with larger

diameter.

Table 4.1. The size of NRs before and after the separation (unit:nm).

L d L/d

Mother 58.54 8.06 7.30

Separated 65.73 7.87 8.35

Through a single centrifugation, the yield of separation was 20 to 60 percent

depending on the property of as-made solution such as the aspect ratio of NRs, the



93

fraction of nanosphere and the diameter difference between NRs and nanospheres.

The yield can be increased up to 70 to 80 percent by repeated centrifugation of the

supernatant solution containing NPs which remain after the previous centrifugation.

But more than 3 times of centrifugation results in the irreversible precipitation of NRs

due to the loss of stabilizer on the surface of the particles resulting from the abrasion

of the particles along the wall of the centrifuge tube18 .

4.4.2. Separation of different aspect ratios.

The NR solution separated from spherical particles by the centrifugation above

was further centrifuged. The centrifuge was stopped every 10 minute and the

deposit at the bottom was taken.

Figure 4.7 shows the UV-Vis-NIR spectra of the centrifuged sample as a

function of time. By the red-shift of LSP and blue-shift of TSP, it is apparent that the

aspect ratio of the NRs obtained at different time becomes larger as a function of time

indicating longer NRs sediment later then the shorter NRs. Also the shape of TSP

changes as a function of time: the solution obtained earlier has a broad shoulder and it

is suppressed towards the later stage.

In our synthesis, generally longer NRs have smaller diameter (Figure 4.8).

Our result shows that relatively shorter NRs having larger diameter sedimented faster

which agrees with the theoretical explanation that the diameter is the key factor for



94

the sedimentation of NRs.

400 600 800 1000 1200

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8


Wavelength(nm)

30min

20min

10min

Figure 4.7. The UV-Vis-NIR spectra of the centrifuged samples as a function of time.

20 40 60 80 100 120 140 160

4

5

6

7

8

9

10

11

L e n g t h ( n m )

Diameter(nm)

Figure 4.8. The plot of relationship between diameter and length of the NRs.

4.4.3. Effect of the surfactant concentration

Centrifugation has been routinely done to remove excess stabilizer for the



95

preparation of TEM grids. During many trials, we have found that there is a thin

layer of dark deposition spread on the side wall of the centrifuge tube. We rinsed the

wall with distilled water and obtained pure NR solution. This is observed only when

as-made solution is centrifuged. When we ran 2nd

round of centrifugation to increase

the yield of separation or to remove more surfactant, and hence distilled water is

added to the concentrated solution, this film was not observed. As it was pointed in

chapter 2, we used higher concentration of surfactant to obtain NRs with higher

aspect ratio and lower byproduct. The concentration of surfactant is almost double

the concentration used in most literature. We deduced from this unique observation

that the concentration of surfactant may affect the parameters of the centrifugation.

We set out an experiment where we divided as-made solution into different

tubes and added different amount of CTAB. We prepared 4 tubes which contain 40

ml of as-made solution. 0.3, 0.9 and 1.2 g of CTAB were added to each tube.

They were centrifuged at 5600 G for 30min. After the centrifugation, concentrated

bottom solutions were removed by a syringe and the UV-Vis-NIR spectra were taken

without dilution. The supernatants were poured out and 5ml of distilled water was

added to each tube to rinse the deposit on the side wall.

Figure 4.9 shows the UV-Vis-NIR spectra of the bottom solutions and the

solutions from the depositions taken from 4 different tubes. As CTAB concentration



96

was increased, the intensity of bottom solution was decreased while the intensity from

the deposition on the side wall was increased. Decrease in the intensity of the bottom

solution indicates the slower sedimentation of spherical NPs.

Figure 4.9. UV–Vis-NIR spectra of the separated solutions. a) Deposit on the

bottom and b) deposit on the side wall as a function of the amount of extra CTAB

added(g/50ml).

a)

400 600 800 1000 1200

0.0

0.1

0.2

O p t i c a l d

e n s i t y

Wavelength(nm)

0

0.3

0.6

0.9

1.2

b)

400 600 800 1000 1200

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7


Wavelength(nm)

0

0.3

0.6

0.9

1.2



97

The yield of obtaining NRs from different conditions was shown in Figure

4.10. When extra CTAB was added, the yield of separation was increased from 36

to 74% as shown from the spectra.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 0.5 1 1.5

the amount of CTAB added (g)

Y i e l d

o f s e p a r a t i o n

Figure 4.10. The yield of obtaining NRs as a function of amount of CTAB

contained in the solution

Therefore the addition of surfactant prevents the sedimentation of nanosphere

and promotes the deposition of NRs on the side wall of the centrifuge tubes.

We need to understand why nanospheres do not sediment fast with increased

surfactant concentration and how the efficient deposition of NRs on the side wall is

promoted. First, the viscosity of the solution changes upon adding surfactant and

affects the sedimentation velocity of the particles19. But this can only partially

explain the slower sedimentation of nanosphere. We still need to explain why rods



99

4.4.4. Separation by flocculation of rods in gravitational field

Jana13

reported surfactant assisted ordering of NRs in a concentrated

dispersion which results in the phase separation of NRs from the spheres. It is

observed that NRs precipitated at the bottom of the flask while spheres stayed in the

supernatant. This observation was attributed to the higher concentration of particles

and addition of extra surfactant which assists the formation of liquid crystal of NRs.

But the mechanism of promoting the liquid crystalline phase of NRs by surfactant as

well as the specific concentration of NRs at which LC phase starts to form were not

mentioned.

We tried to use the same experimental parameters to reproduce the result of

Jana’s. However we couldn’t observe any separation.

It is worth to notice that the size of NRs which Jana used was larger since they

were made by sequential growth method 8. It was mentioned that the width was

12nm and the aspect ratio was 20, hence the length of NRs was 240nm.

We reasoned that the size of NRs might be the key parameter for this

separation. When we used the NRs of larger size (average length 80 nm and average

diameter 11 nm) which were synthesized by reinitiating reduction of as-made NRs

solution (Chapter 2), we also observed the similar phenomena without concentrating

the solution and adding extra surfactant.



100

When the as-made solution was observed after 24hrs, a thin layer of dark color

was formed at the bottom of flask, which was so stable that it didn’t flow when the

supernatant was poured out. Distilled water was added and the deposition was

redispersed.

Figure 4.11. Gold NRs flocculation observed under the microscope with crossed

polarizers. The scale bar is 100µm.

Figure 4.11 shows a cross polarized microscopy image of the solution in

which the precipitation was redispersed. We observed gold NRs flocculation in

micron size showed birefringence. The size of theses domains is about 1 µm in

diameter. Since the gold NRs form flocculation, we sonicated the solution for 1hr to

ensure the deflocculation of gold NRs and took the UV-Vis-NIR spectrum of the

solution.

Figure 4.12 shows the UV-Vis-NIR spectra of solution before and after

separation. The solution from the precipitate has very small transverse peak and most

of the shoulder was removed indicating that it contained mostly longer NRs.



101

This result cannot be explained by the Jana’s explanation since the

concentration of gold NRs in the solution is extremely low.

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

350 550 750 950 1150 1350

Wavelength(nm)

O p

t i c a l d e n s i t y

Precipitation

Mother

Supernatant

Figure 4.12. UV–Vis-NIR spectrum of separated solutions in gravitational field.

Since the stationary-state sedimentation velocity is proportional to the square

of the radius of the particle, theses flocculations sediment approximately 10000 times

faster than individual nanoparticles14

. Therefore we believe that NRs of large aspect

ratio form flocculation which has ordered structure and eventually sediment in

relatively short time. However, the mechanism of forming flocculation needs to be

studied.



102

In the general literature for centrifuge 22 , the swinging bucket rotor is

recommended if the purpose of centrifugation is to separate molecules or organelles

on the basis of their movements through a viscous field. In the case of fixed angle

rotor, the sediment is forced against the side of the centrifuge tube, and then slides

down the wall of the tube which leads to abrasion of the particles along the wall of the

centrifuge tube. We observed the irreversible aggregation of gold NRs from the

repeated centrifugation in our own experiment and in the literature. Even though the

type of rotor was rarely mentioned in the literature, since the rpm is usually higher

than the maximum rpm to be obtained by swing-bucket rotor, we assumed that they

all used a fixed angle rotor which causes such a problem.

4.5. Conclusions

In conclusion, we elucidate the hydrodynamic behavior of nanoparticles of

various shapes to identify the parameters to achieve efficient separation. For a given

L/d, whether the rods or spheres sediment faster is mainly controlled by the relative

diameters of the particles. By centrifugation, we succeeded in separating NRs from

spherical nanoparticles and to separate different aspect ratio of NRs. The purity of

the NRs is more than 99% (number fraction) and the yield of separation through a



103

round of centrifugation can be increased up to 78% by increasing CTAB

concentration which affects the colloidal interaction.

Centrifugation can be a choice of separation method when the size and shape of

the mixed particle result in sufficiently different sedimentation coefficient between

different particles.



104

References

1 (a) Burda,C.; Chen,X.; Narayanan, R.; El-Sayed, M. A., “Chemistry and

properties of nanocrystals of different shapes” Chemical Reviews 2005, 105,

1025 (b) El-Sayed, M.A., “Some Interesting Properties of Metals Confined inTime and Nanometer Space of Different Shapes” Acc.Chem. Res. 2001, 34, 257 (c)

Schmid, G.; Corain, B. “Nanoparticulated Gold: Syntheses, Structures, Electronics,

and Reactivities” European Journal of Inorganic Chemistry 2003, 17 , 3081 (d)

Daniel, M.; Astruc, D. “Gold Nanoparticles: Assembly, Supramolecular

Chemistry,Quantum-Size-Related Properties, and Applications toward Biology,

Catalysis,and Nanotechnology” Chemical Reviews 2004, 104, 293.

2

(a) Yin,Y.; Alivisatos,A.P. “Colloidal nanocrystal synthesis and the organic– inorganic interface” Nature 2005, 437 , 664 (b) Lisiecki I. “Size, Shape, and

Structural Control of Metallic Nanocrystals” J. Phys. Chem. B 2005, 109,

12231.

3 Murray, C. B.; Norris, D. J.; Bawendi, M. G. “Synthesis and characterization of

nearly monodisperse CdE (E = sulfur, selenium, tellurium) semiconductor

nanocrystallites” J. Am. Chem. Soc. 1993, 115, 8706.

4 Whetten, R. L.; Khoury, J. T.; Alvarez, M. M.; Murthy, S; Vezmar, I.; Wang, Z. L.;

Stephens, P.W.; Cleveland, C. L.; Luedtke, W. D.; Landman,U. “Nanocrystal gold

molecules” Adv. Mater . 1996, 8 , 428.

5 Liua,F.; Koa,F.; Huang,P.,Wub, C.; Chuc, T. “Studying the size/shape separation

and optical properties of silver nanoparticles by capillary electrophoresis” Journal

of Chromatography A, 2005,1062, 139.

6 Wei, G-T; Liu, F-K; Wang, C. R. Chris. “Shape separation of nanometer cold

particles by size-exclusion chromotography” Analytical Chemistry 1999, 71,

2085.

7 Yu,Y; Chang,S; Lee,C; and Wang, C. R. C. “Gold Nanorods: Electrochemical

Synthesis and Optical Properties” J. Phys. Chem. B, 1997, 101, 6661.

8 Jana,N. R; Gearheart, L,; Murphy,C. J. “Wet Chemical Synthesis of High Aspect

Ratio Cylindrical Gold Nanorods” J. Phys. Chem. B, 2001, 105, 4065.



105


Nanorods(NRs) Using Seed-Mediated Growth Method” Chem.Mater . 2003, 15,

1957.



Reviews 2005, 249, 1870.

11 Kim, F.; Song, J. H.;Yang, P.,”Photochemical synthesis of gold nanorods” J. Am.

Chem. Soc., 2002,124, 14316.

12

Nikoobakht,B. “Synthesis, characterization and self-assembly of gold nanorods andsurface-enhanced Raman studies” Ph.D thesis 2002 Georgia Institute of Technology.

13 Jana, N. R. “Nanorod shape separation using surfactant assisted self assembly”

Chem. Comm. 2003,1950.

14 Hiemenz, P.C.; Rajagopalan, R. “Pinciples of Colloid and Surface Chemistry” 3rd

ed., Marcel Dekker 1997.

15 Ramaswamy, S. “Issues in the statistical mechanics of steady sedimentation”, Adv.

Phys. 2001,50, 297.

16 Happel, J.; Brenner, H., “ Low Reynold Number Hydrodynamics”, 2

nd ed.

Noordhoff International Publishing, Leyden 1973.

17 de la Torre, J.G.; V. A. Bloomfield,V.A. “Hydrodynamic properties of complex,

rigid, biological macromolecules-theory and applications” Q. Rev. Biophys. 1981,14,

81.

18 Mikkelsen, S.R.; Cortón E., “Bioanalytical chemistry” John Wiley & Sons, 2004.

19

η

ρ−ρ=

g)(R

9

2v 12

2

s When η increased, v decreased.

20 Everett, D.H. “Definitions, Terminology and Symbols in Colloid and Surface

Chemistry - Part I, Appendix to Manual of Symbol and Terminology forPhysicochemical Quantities and Units” Pure Appl. Chem., 1972, 31, 579.



107

5. Chapter V

The Optical Properties of Gold Nanorods

Abstract

The current intense interest in metal nanoparticles for their applications in

optical sensors and biological imaging is mainly based on their optical properties. In

this study, the optical absorption of gold nanorods (NRs) was investigated. The

theoretical calculation and experimental results were compared and the color of gold

NRs was systematically identified for a wide range of aspect ratios. We also

investigated the effect of medium refractive index on the absorbance of gold NRs,

which would lead us to the proper choice of solvents for better application of gold

NRs.



108

5.1. Introduction

Although bustling studies on gold nanoparticles have led to dramatically

increasing number of publications, especially in the context of emerging nanoscience

and nanotechnology, colloidal gold has been known for centuries as a colorant to

make ruby glass and for coloring ceramics. The brilliant color is due to the presence

of the metallic gold in colloidal form, although this fact was not clearly known until

Michael Faraday’ lecture on the “Experimental Relations of Gold (and other metals)

to Light”1.

A good understanding of the optical properties of gold nanoparticles is

important for several reasons. The optical properties are directly correlated to size

and shape of the particles. Hence, the formation of NPs in solution can be easily

monitored using UV–Vis–NIR spectroscopy. Also, optical properties may be very

sensitive to the surface processes such as adsorption, precipitation and electron

transfer. The fact that the spectral characteristics of gold nanoparticles are quite

sensitive to their surroundings enables them to be used as sensors. Gold

nanoparticles have been used in recent years as markers for optical detection of

biospecific intermolecular interactions2.

The optical properties of noble metal particles originate from surface

plasmons. These phenomena occur when electromagnetic field interacts with



109

conduction band electrons and induces the coherent oscillation of the free electrons.

As a result, a strong absorption band appears in some region of the electromagnetic

spectrum depending on the size of the particle3. This plasmon absorption is a small

particle effect. They are absent in the individual atoms as well as in the bulk. Mie

first explained this phenomenon theoretically by solving Maxwell’s equations for a

radiation field interacting with a spherical metal particle under the appropriate

boundary conditions4.

Scheme 5.1. Schematic of plasmon oscillation for a metal sphere5.

For gold NRs, the plasmon absorption splits into two bands corresponding

to the oscillation of the free electrons along and perpendicular to the long axis of the

rods. The transverse mode (transverse surface plasmon peak: TSP) shows a

resonance at about 520 nm, while the resonance of the longitudinal mode

(longitudinal surface plasmon peak: LSP) strongly depends on the aspect ratio of

NRs6 . As the aspect ratio is increased, the longitudinal peak is red-shifted.

To account for the optical properties of NRs, it has been common to treat them



110

as ellipsoids, which allows the Gans formula (extension of Mie theory) to be applied.

A geometrical factor P j was introduced to calculate the absorbance of light by

ellipsoids. Gans’ formula7 for randomly oriented elongated ellipsoids in the

dipole approximation is

∑=

+⎥⎥

⎦

⎤

⎢⎢

⎣

⎡

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛ −+

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

=C

A j

j

j

j

p

P

P

P

V N 2

2

2

1

2

22/3

1

1

3

2

ε ε ε

ε

λ

πε γ

α

α

where N p represents the number concentration of particles, V the single particle

volume, λ the wavelength of light in vacuum, ε α the dielectric constant of the

surrounding medium, and ε 1 and ε 2 are the real (n2 - k 2) and imaginary (2nk ) parts of

the complex refractive index of the particles. The geometrical factors P j for

elongated ellipsoids along the A and B/C axes are respectively given by

( )

2/1

2

22

2

2

2/1

11

1ln

2

11

⎟⎟ ⎠ ⎞⎜⎜

⎝ ⎛ −=

−==

⎥⎦

⎤⎢⎣

⎡−⎟

⎠

⎞⎜⎝

⎛

−

+−=

Ld Le

PPP

e

e

ee

eP

AC B

A

Figure 5.1 shows the absorbance spectra calculated using the above

expressions. The optical parameters for bulk gold are taken from the literature8. The

refractive index of the medium was assumed to be constant and equal to 1.333 for

water. The maximum of the longitudinal absorbance band shifts to longer



111

wavelengths with increasing aspect ratio. There is a small shift of the transverse

resonance maximum to shorter wavelengths with increasing aspect ratio.

Electron microscopy reveals that most NRs are more like cylinders or sphero-

capped cylinders than ellipsoids. However, an analytical solution for such shapes does

not exist.

Figure 5.1. Absorbance spectra calculated with the expressions of Gans for elongated

ellipsoids using the bulk optical data for gold. a) The numbers on the spectral curves

indicate the aspect ratio ( L/d ). b) Enlargement of the shaded area of a) showing slight

blue shift of TSP on increasing aspect ratio.

The discrete dipole approximation (DDA) method is a numerical method first

introduced by Purcell and Pennypacker 9 in which the object studied is represented as

a cubic lattice of N polarizable point dipoles localized at r i, i = 1, 2, ... , N , each one

characterized by a polarizability α i. There is no restriction on the localization of cubic

lattice sites so that the DDA represents a particle of arbitrary shape and composition.

a) b)

400 600 800 1000 1200 1400

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2


Wavelength(nm)

Aspect ratio 3 to 10

440 460 480 500 520 540 550

0.00

0.01

0.02

0.03

0.04

0.05

0.06


Wavelength(nm)

Aspect ratio 3 to 10



112

Recently, Schatz and co-workers showed that DDA is suited for optical

calculations of metallic systems with different geometries and environments through

the extensive studies on the extinction spectrum and the local electric field

distribution in metal particles5.

Brioude et al.10 used DDA method to simulate the absorption efficiency of

gold NRs where they took into account the real shape of gold NRs. A dominant

surface plasma band corresponding to the longitudinal resonance is observed. Its

maximum position shifts to the red as the aspect ratio increases. These data are in

good agreement with previous theoretical calculations based on classical electrostatic

predictions which assume that gold NRs behave as ellipsoidal particles6. From the

experimental point of view, good agreement with the published data for gold NRs is

obtained.

The color of colloidal gold depends on both the size and shape of the particles,

as well as the refractive index of the surrounding medium11. Understanding the color

change of gold NRs surrounded by different medium is quite important. So far, gold

NRs have been made in aqueous solution. In the mean time, surface modification of

gold NRs for further application often involves using a mixture of organic solvents

due to the limited solubility of the reactant in water. Since, most characterization is



113

done by UV-Vis-NIR spectrum and TEM, the color change of NRs solely by the

change of medium has to be studied before advancing to modification of NRs.

A number of groups have reported the absorption spectra of gold spheres

prepared in solvents other than water and they were compared with theoretical

calculations12. For gold NRs, theoretical prediction of their optical properties of gold

NRs upon changing medium was made by several groups and compared with the

experimental results13,14,15. They found that there is a linear relationship between

LSP and dielectric constant of medium. But so far these studies were limited in the

narrow range of aspect ratios. Also the process to transfer gold NRs to organic

solvent is non-trivial.

The purpose of this section is to have comprehensive knowledge of the optical

properties of gold NRs. We compared the optical properties of gold NRs from the

theoretical calculation and experimental results for a wide range of aspect ratio of

gold NRs and systematically identified the color of gold NRs as a function of aspect

ratio. We also investigated the effect of medium refractive index on the color of gold

NRs and converted the spectral data to conform with the language of color science16.

5.2. Experimental




114


The changes in the absorption spectrum of solution of NRs were determined at

different solvents/water ratios using glycerol, dimethylsulfoxide (DMSO),

dimethylformamid (DMF) and acetonitrile. Water was purified by ion exchange. The

sample was immersed in a thermostat bath at 25°C and the absorption spectrum of the

sample was measured when the sample solution was in equilibrium with the

thermostat temperature. UV-Vis-NIR spectra were acquired with a Cary 5G

spectrophotometer. Morphology and mean size of NPs were examined by TEM

(JEOL100 at 100 KV).

To simulate the color of gold NRs from the calculated absorption spectra and

experimental spectra, CIE XYZ tristimulus values were calculated 16. The

absorbance ( A) was converted to transmittance (T ) according to the following

equation.

T

1log A 10=

CIE XYZ tristimulus values were calculated by the integration of the

transmittance values T(λ ), the relative spectral energy distributions of the illuminant

E (λ), and the standard observer functions x(λ ), y(λ ), and z(λ ). We use the spectral

distribution power function of daylight D65. The integration is approximated by

summation, thus:



115

∫= )( x)( E )(T k

1 X λ λ λ

∫= )( y)( E )(T k

1Y λ λ λ

∫= )( z)( E )(T k

1 Z λ λ λ

where ∫= )( y)( E k λ λ and λ = wavelength.

The amounts of red, green, and blue needed to form any particular color are

called the tristimulus values and are denoted X, Y, and Z, respectively. A color is then

specified by its trichromatic coefficients, defined as

x = X / ( X + Y + Z )

y = Y / ( X + Y + Z )

z = Z / ( X + Y + Z )

x + y + z = 1.

The color was identified by positioning x and y values in the CIE chromaticity

diagram.


5.3.1. Color of gold nanorods in aqueous solution.

The relation between aspect ratio and LSP is plotted according to Gans’

calculation6 and DDA simulation5 and compared with experimental results. In both



116

theoretical data, the dielectric constant of 1.77 was used which corresponds to the

value for water where the NRs are dissolved.

Calculation by Gans’ theory and simulation by DDA are plotted in Figure 5.2

showing the similar trend where LSP lineary increases with increasing aspect ratio.

There is a small discrepancy in the LSP between two predictions which is within

20nm originating from the initial assumption of different shape of NRs (ellipsoid for

Gan’s calculation and spherocyliner for DDA). Experimental results are overlaid

showing good agreement with the predicted values. DDA simulation has better

agreement than Gans’ calculation. From these results, it can be said that the use of

dielectric constant of water is reasonable to predict the optical properties of gold NRs.

There is a discussion about the effect of capping material on the optical

properties of NRs in a way that CTAB and the capping structure might be involved in

determining the dielectric constant of medium15. Since NRs are stabilized by

surfactant, we cannot exclude the effect of surfactant. To probe the effect of

capping materials on the optical properties of gold NRs, the surfactant in the gold NR

solution was removed as much as possible by centrifugation. The concentrated

solution was diluted and different amount of CTAB was added which will affect the

structure of CTAB on the surface of gold NRs. We found that there is no

dependence of the concentration of CTAB on the optical properties of gold NRs



117

within the concentration range studied. Therefore the capping material has no

significant role in optical properties of gold NRs while the refractive index of shell is

important in core shell structure of nanomaterial17.

600

800

1000

1200

1400

1600

1800

2000

2200

1 3 5 7 9 11 13 15 17

aspect ratio

L S P ( n m )

Figure 5.2. Longitudinal surface plasmon peak versus the aspect ratio of gold NRs.

Simulation results using the DDA method and the corresponding fit (solid line) and

Gans’ calculation (dotted line). Experimental data from the work of van der Zande et

al18(square). Experimental data from our study (triangle).

It is often mentioned in the literature that the optical properties of gold NRs

are tunable as a function of NP size and shape19. But the color has not been defined

systemically as a function of aspect ratio.

One has to consider the fact that the TSP is not sensitive to the change of

aspect ratio and only the LSP is sensitive to the aspect ratio. This visible light region

consists of a spectrum of wavelengths, which range from approximately 700 nm to

400 nm. Therefore once the LSP goes beyond 700nm, basically the color of gold NRs



118

solution has to be the same. From the simulation and our experimental result, LSP

of 700 nm corresponds to the NRs of aspect ratio around 3.

Therefore the color change could be only observed for relatively short range of

aspect ratios. We simulated the color of the gold NRs using theoretical absorbance

data. As predicted, the color does not change significantly when L/d is greater than

4 (Figure 5.3).

Figure 5.3. The simulated color of gold NR solution of different aspect ratios and their

trichromatic coefficients.

Figure 5.4 shows the UV-Vis-NIR spectrum of gold NR solution

experimentally prepared. Each solution contains mostly NRs as it can be seen from

the small TSP in the spectrum. We obtained these pure NR solutions by optimizing

synthesis condition and separating byproduct from the rods by centrifugation (for

details, see chapter 2 and 4). The transverse peak matches with theoretical

prediction where TSP shifts to shorter wavelength as increasing aspect ratio.



119

Figure 5.4. a) UV-Vis-NIR spectra of gold NR solutions having different aspect

ratios and b) transverse peak in detail.

Figure 5.5 shows the photograph of the gold NR solution and the color

patches simulated using theoretical absorbance data equivalent to the aspect ratio of

a)

400500 1000 1500 2000

0.0

0.1

0.2

0.3

0.4

0.5


Wavelength(nm)

Aspect ratio

3.0

4.4

5.6

7.4

14.5

b)

400 420 440 460 480 500 520 540 560 580 600

0.00

0.02

0.04

0.06

0.08

0.10


Wavelength(nm)

Aspect ratio

3.0

4.4

5.6

7.4

14.5



120

gold NRs. The color of solution is basically the same beyond the aspect ratio around

4. Also the color of the gold NR solution we made is identical to the simulated color

(Figure 5.5 b).

Therefore in a visible region, the dramatic color change can not be achieved

only by changing aspect ratio. But the tunability of optical properties of gold NRs as

a function of the aspect ratio provides potentials to use gold NRs as an optical filter in

near infrared region.

Figure 5.5. a) Photograph of 4 solutions of colloidal gold prepared in water. Aspect

ratios are 2.6, 4.1, 5.6, 7.4 (from the left), respectively. b) The simulated color of gold

NR solution of different aspect ratios.

We found that the color in a visible region is rather sensitive to the amount of

spherical particles included as byproduct since surface plasmon peak of sphere

positions between 500 to 550nm.



121

Figure 5.6 shows the simulated color of gold NR (aspect ratio is 5) solution

having different amount spheres (surface plasmon peak at 520nm) as byproduct. The

color changes from purple to brown as the amount of byproduct decreases.

Figure 5.6. The color of gold NR solution of having different amount of spheres as byproduct. a) 50%, b) 30%, c) 10%, and d) 0%.

5.3.2. Effect of refractive index of medium

We simulated dependence of absorbance of gold NR solution on changing

refractive index of medium. The LSP increases linearly as the refractive index of

medium increases (Figure 5.7). And the effect of medium is pronounced for longer

NRs as is shown from the slop of the plot.



122

1.32 1.34 1.36 1.38 1.40 1.42 1.44

700

800

900

1000

1100

1200

1300

L S P ( n m )

Refractive index

Aspect ratio

8

6

4

3

Figure 5.7. Calculated LSP as a function of refractive index of medium.

To validate the predicted LSP and the experimental value, we conduct the

study on the changing refractive index of medium by changing the volume fraction of

water and glycerol. We choose glycerol because it does not cause precipitation, and

there is no need to exchange the capping material to transfer gold NRs to different

medium.

Figure 5.8 shows the experimental data for the change of LSP upon adding

glycerol to water. As we expected, LSP red-shifts as the medium refractive index

increases.



123

0

0.05

0.1

0.15

0.2

0.25

350 450 550 650 750 850 950

Wavelength(nm)


Increasing n

Figure 5.8. Absorption spectra of gold NRs with aspect ratio L/d = 3.5 in various

glycerol/ water ratios. From left, water, 20, 40 and 50 % (v/v) of aqueous glycerol

solutions.

Using the refractive index of the mixture (shown in Table 5.1) in the Gans’

equation we calculated the absorbance of NR solutions, and compared the calculations

to the experimental values obtained (Figure 5.9). The experimental results show

good agreement with the calculated values. The effect of medium is indeed

pronounced for longer NRs.



124

700

750

800

850

900

950

1000

1050

1100

1150

1200

1.3 1.35 1.4 1.45

The refractive index of medium

L S P ( n m )

L/d = 3.5

L/d = 7.5


refractive index of medium (water/glycerol mixture). Filled dots represent the

experimental data and straight line is calculated data.

Table 5.1. The refractive index of water/glycerol mixture.20

Volume fraction Refractive index0% glycerol 1.46

10% glycerol 1.33

30% glycerol 1.34

50% glycerol 1.37

70% glycerol 1.40

90% glycerol 1.43

There is a great deal of multidisciplinary research in recent years on the

assembling NPs using linkers. Using organic solvent is inevitable since most of the

linker molecules are water insoluble. It is often the cases that the aqueous gold NR

solution is mixed with organic solvent. Depending on the mixing ratio, gold NRs

may aggregate and precipitate. Therefore, it is important to find the optimum mixing



126

In both cases, there is an increasing deviation between theoretical LSP and

experimental LSP as the fraction of DMSO increases. Finally LSP starts to

decreases, which indicates the aggregation of NRs. In the case of mixing ratio of 1

to 1, NRs precipitate in several hours.

DMF shows results that are similar to DMSO. Since DMF has a higher

refractive index than that of water, as the fraction of DMF increases, the LSP

increases. When the fraction of DMF is increased beyond 50%, NRs start to

aggregate and precipitate.

Acetonitrile is among the solvents used for preparing self assembly of gold

NRs21. It is reported that the spectral properties of gold NRs were unaffected upon

increasing the composition of acetonitrile (acetonitrile: water = 4:1) and the rods were

quite stable for several hours. Acetonitrile could be a good choice of solvent since it

has similar refractive index (n=1.3441) as water. When acetonitrile was added in

minor fraction to gold NR solution, there is indeed no change in absorbance peaks.

But as the fraction of acetonitrile increases, LSP red shifted in a large amount with

decreasing intensity indicating the aggregation of NRs. Figure 5.11 shows the UV

spectrum of gold NRs in aqueous solution and solvent mixture (acetonitrile: water =

4:1 by volume. This ratio is used in the literature). We found that the gold NRs in

this solvent mixture precipitate out the next day.



128

5.4. Conclusions

In summary, we found that both Gans’ theory and DDA simulation provide a

good representation of the experimental results for absorbance of gold NRs. The

color of gold NR aqueous solutions matches well with the simulated color which is

generated using theoretical absorbance of equivalent aspect ratio. The medium

refractive index affects the absorbance of gold NRs. There is a linear relationship

between LSP and refractive index. . The effect of the medium is greater for longer

NRs.

Our results suggest that the knowledge of the refractive index of a known

solvent and absorbance of gold NRs provide an easy method to determine the aspect

ratio of the NRs. Also, our results will be useful as the foundation for the potential

optical applications of gold NRs.



129

References

1 Faraday, M., “Experimental Relations of Gold (and other metals) to Light” Philos.

Trans. 1857, 147 , 145.

2. a) Huang, X; El-Sayed, I. H.; Qian, W; El-Sayed, M. A.“Cancer Cell Imaging and

Photothermal Therapy in the Near-Infrared Region by Using Gold Nanorods” J.

Am. Chem. Soc . 2006, 128, 2115, b)Englebienne P.; Van Hoonacker, A;

Verhas, M. “Surface plasmon resonance: principles, methods and applications in

biomedical sciences” Spectroscopy-an international journal, 2003, 17 , 255.

3 Kreibig ,U; Vollmer,M. “Optical Properties of Metal Clusters” Springer, 1995.

4 Mie,G. “Contributions to the Optics of Turbid Media, Especially Colloidal Metal

Solutions” Ann. Physik 1908, 25, 377.

5 Kelly, K. L.; Coronado, E.; Zhao, L. L.; Schatz, G. C. “The optical properties of

metal nanoparticles: The influence of size, shape, and dielectric environment” J.

Phys.Chem. B. 2003, 107 , 668.

6 Link, S; El-Sayed M.A. “Spectral properties and relaxation dynamics of surface

plasmon electronic oscillations in gold and silver nanodots and nanorods” J. Phys.

Chem B 1999, 103, 8410.

7 a) Gans,R. “The Form of Ultramicroscopic Gold Particles” Ann. Physik 1912, 37 ,

881 b) Gans,R. “Form of ultramicroscopic particles of silver” Ann. Physik 1915, 47,

270.

8 Johnson, P. B.; Christy, R. W. “Optical Constants of the Noble Metals” Phys. Rev.

B, 1972, 6 , 4370.

9 Purcell, E. M.; Pennypacker, C. R. “Scattering and absorption of light by

nonspherical dielectric grains” Astrophys. J . 1973, 186, 705.

10 Brioude, A.; Jiang,X. C.; Pileni, M. P. “Optical Properties of Gold Nanorods: DDA

Simulations Supported by Experiments” J. Phys. Chem. B 2005, 109, 13138.



130

11 Khlebtsov, N.G.; Trachuk,L. A.;. Mel’nikov, A. G. “The Effect of the Size, Shape,

and Structure of Metal Nanoparticles on the Dependence of Their Optical Properties

on the Refractive Index of a Disperse Medium” Optics and Spectroscopy, 2005,98 ,

77.

12 Underwood, S.; Mulvaney, P. “Effect of the Solution Refractive Index on the

Color of Gold Colloids” Langmuir 1994, 10, 3427.

13 Link,S.; Mohamed, M. B.; El-Sayed M. A. “Simulation of the Optical absorption

Spectra of Gold Nanorods as a Function of Their Aspect Ratio and the Effect of the

Medium Dielectric Constant” J. Phys. Chem. B 1999, 103, 3073.

14 Yang, J.; Wu, J.; Wu, Y.; Wang, J.; Chen, C. “Organic solvent dependence of

plasma resonance of gold nanorods : a simple relationship” Chemical Physics

Letters 2005, 416 , 215.

15 Al-Sherbini, A. “Thermal instability of gold nanorods in micellar solution of

water/glycerol mixtures” Colloids and Surfaces A: Physicochem. Eng. Aspects 2004,

246 , 61.

16 Nassau, K. “Color for Science, Art and Technology” Elsevier Science &

Technology Books 1998



Reviews 2005, 249, 1870

18 van der Zande, B.M.I.; Böhmer, M.R.; Fokkink,L.G.J.; Schönenberger, C.

“Colloidal Dispersions of Gold Rods: Synthesis and Optical Properties” Langmuir

2000, 16, 451.19 Murphy,C.J.; Sau,T.K.; Gole, A.M.; Orendorff, C.J; Gao,J.; Gou,L.; Hunyadi, S.E.;

Li,T. “Anisotropic Metal Nanoparticles: Synthesis, Assembly, and Optical

Applications” J. Phys. Chem. B 2005, 109, 13857.

20 Borst, J.; Hink, M. A.; Hoek, A.; Visser, A. J. W. G. “Effects of refractive index and

viscosity on fluorescence and anisotropy decays of enhanced cyan and yellow

fluorescent proteins” Journal of Fluorescence 2005, 15, 153.



131

21 Joseph, S. T. S.; Ipe, B. I.; Pramod, P.;Thomas, K.G. “Gold Nanorods to

Nanochains: Mechanistic Investigations on Their Longitudinal Assembly Using α-

ϖ-Alkanedithiols and Interplasmon Coupling” J. Phys. Chem. B 2006, 110, 150.



132

6. Chapter VI

Optical Properties of PVA/Gold Nanorods nanocomposites

Abstract

Gold nanorods (NRs) with well defined aspect ratio were introduced into a

poly (vinyl alcohol) (PVA) matrix by means of solution-casting techniques. The

resulting nanocomposites film was drawn to induce the uniaxial alignment of gold

NRs. The absorption spectra of the drawn nanocomposites in the UV-Vis-NIR

wavelength region were found to depend strongly on the polarization direction of the

incident light. The observed optical anisotropy originates from uniaxially oriented

gold NRs. It is shown that the polymer-inorganic nanocomposites can be employed

as color polarizing filters. Films of different colors were prepared by changing aspect

ratio and mixing rod and spheres, which also show the sensitive color change upon

the changing polarization angle. By using higher aspect ratio of NRs (L/d=6.3 and

above), the polarizing filters in the NIR region was achieved.



133

6.1. Introduction

Because the plasmon resonance peaks of gold nanorods (NRs) arise from the

anisotropy of NRs, the optical property of a single NR is greatly influenced by the

polarization of incident light1. When the incident light is polarized parallel to the

main axis of the particle, the spectrum displays only the longitudinal resonance. In the

case of light polarized perpendicular to the main axis of the rod-shaped particles, the

spectrum displays only the transverse resonance. The UV-vis spectrum of gold NRs

solution shows the optical response of randomly distributed gold NRs to the

unpolarized incident light. Therefore if NRs have the orientational order by

alignment process and the direction of polarization of incident light is controlled, the

magnitude of both the longitudinal and transverse resonance will be sensitive to the

angle between the orientation of NRs and the polarization of incident light and show

the optical dichroism in a certain wavelength region.

Obtaining the alignment by stretching the film has the potential for many

applications since there is increasing interest to make polymer nanomaterial

composite where the optical properties of nanoparticles can be utilized to make some

smart applications in the fields of sensing, optics, or electronics2. Immobilization of

NRs on the surfaces or their incorporation into bulk materials is essential and the

optical properties of nanocomposites with the same nanoparticles are subject to



134

change depending on the polymer matrix.

Al-Rawashdel et al. prepared polyethylene/gold rods nanocomposites and

found that the composite films exhibit linear dichroic properties which result from a

net orientation of the rods3 . The plasmon resonance absorption maxima occur at

the longest wavelengths when the incident electric field is polarized along the

direction of particle orientation, and are blue-shifted as the polarization angle is

perpendicular to the orientation direction. These results were confirmed to be in

qualitative accord with Rayleigh scattering theory and T-matrix scattering

calculations4.

Dirix et al. utilized these properties and showed that the polymer-inorganic

nanocomposites can be employed as color polarizing filters5. They incorporated

alkanethiol coated spherical gold particles into a polyethylene matrix. Upon uniaxial

drawing, spherical shapes of individual nanoparticles form aggregation of different

aspect ratio. The absorption spectra of the drawn nanocomposites in the visible

region were found to depend strongly on the polarization direction of incident light

(demonstrated in Figure 6.1).



135

Figure 6.1. a) UV-Vis spectra of drawn nanocomposites comprising of high-density

polyethylene and gold, taken in linearly polarized light at different angles between the

polarization direction of the incident light and the drawing direction. b) Twisted-

nematic liquid crystal displays (LCD) equipped with a drawn polyethylene-silver

nanocomposite. The “M” represents the on state, the drawing direction is in the

picture above parallel and below perpendicularly oriented to the polarizer 6.

The choice of polymer is important because the inorganic particles typically

aggregate in polymer solution owing to their high surface energies which results in

excessive light scattering by the composite material.

Poly (vinyl alcohol) (PVA) offers a suitable matrix for the fabrication of

composites of gold NRs. Synthetic procedures giving NRs in high yield are based

on aqueous solution. Since PVA is water soluble, the nanocomposites can be made

by mixing of the individual components in a polymer solution. When the composite

films are stretched, the gold NRs tend to align in a preferred direction. The

alignment is driven by the mechanical force conveyed through the polymer film.

a) b)



136

Van der Zande et al., prepared thin film of gold NRs dispersed in PVA7.

Although they used PVA, they took somewhat complicated procedure. The gold NRs

were first transferred from water to ethylene glycol and PVA was added to the gold

NRs/ethylene glycol solution. The solution was heated to 195°C and casted on a

glass plate. The resulting film was post processed to insure the quality of film. The

orientation was obtained by stretching film at 120°C. They found that the rods are

completely oriented when the film is stretched 4-6 times its original length. The

polarization absorbance spectra showed polarization dependence. A similar

procedure was adopted by other research groups where nanocomposites are made by

mixing gold NRs aqueous solution directly to PVA aqueous solution. Murphy et al.8

casted solution on a Petri dish and cured for 18h at 80°C. Then the film was

stretched at 60°C by hand and cooled to room temperature. Pérez-Juste et al.9 used

the similar approach to prepare PVA gold NRs composite, but they didn’t specify the

post processes. They briefly mentioned that the NRs were aligned by warming and

stretching (by hand) the composite film. They also mentioned that care should be

taken to avoid overheating, since this can lead to modification of the optical properties

due to thermal reshaping of the rod.

Although several groups show dichroic properties of aligned gold NRs, the

tunablity of peak position was not demonstrated extensively.



137

The aim of this study is to investigate the optical properties of anisotropic gold

particles embedded in PVA films as a function of the net orientation of the rods and

polarization angle. By changing aspect ratio of gold NRs and mixing with spherical

particle, we will probe the potentials to use these nanocomposites as color polarizing

filters. Also, by using higher aspect ratio of NRs, we will probe the possibility that

the optical dichroism can be shifted from the visible to the near IR.

6.2. Experimental



For the preparation of PVA/gold NRs nanocomposites, typically, 10 w % PVA

aqueous solution was prepared by dissolving PVA (Mw: 95000, 98-99% Hydrolyzed,

Aldrich) in 50 ml of water. The solution was stirred at 50°C for 12 hrs.

Concentrated gold NRs solution was mixed with the PVA solution and poured to Petri

dish which is covered for 24 hrs for the formation of uniform thickness of film (the

final concentration of gold NRs in the film is about 6×10-4 wt % . Then the cover was

removed and water was evaporated. After the water evaporation, the polymer film

was carefully peeled from the substrate. The film thickness was approximately

50µm.



138

The film was cut to the size of 1cm × 2cm and mounted on drawing equipment

specially manufactured. The drawing was done on the hot plate at 60°C by moving

the knob on the equipment. The oriented nanocomposites were analyzed in linearly

polarized light at different angles between the polarization direction of the light and

the drawing direction of the nanocomposites.

Absorption spectra were acquired using a CARY 5G UV-vis near-IR

spectrophotometer and UV-vis spectra of composite films were measured with SEE-

1000 microspectrophotometer.

Confocal Scanning Laser Microscopy (CSLM) measurements were performed

in the reflection mode with a Leica TCS NT confocal microscope equipped with a

krypton/argon gas laser and an oil immersion lens.

6.3. Results and discussions

6.3.1. Optical properties of gold NRs in PVA

Figure 6.2 shows the absorbance spectra of randomly oriented gold rods in

water and in PVA films. The films were transparent indicating the colloidal gold rods

were well dispersed in PVA.



139

Figure 6.2. The absorbance spectra of randomly oriented gold rod dispersions. a)

aspect ratio 4.3 , b) aspect ratio 6.3. The blue line is the absorbance spectra of

aqueous gold NR solution. The red line is the absorbance spectra of gold NRs in

PVA film.

a) L/d=4.3

400 600 800 1000

0.0

0.1

0.2

0.3

0.4

0.5


Wavelength(nm)

water

PVA

b) L/d=6.3

400 600 800 1000 1200

0.0

0.1

0.2

0.3


Wavelength(nm)

water

PVA



140

L/d = 4.3

L/d = 6.3

800

850

900

950

1000

1050

1100

1150

1200

1.25 1.3 1.35 1.4 1.45 1.5

Refractive index of medium

L S P

Figure 6.3. Longitudinal surface plasmon peak (nm) versus refractive index of the

medium. Straight line is obtained from the calculation of Gans theory. Diamonds

and triangles are the experimental data.

The spectra of the PVA films were similar to those of the aqueous solution.

The LSP shifted to longer wavelengths due to the change in refractive index of the

medium. This shift can be theoretically predicted by calculating the peak positions

using the expressions of Gans. The optical parameters for bulk gold are taken from

the reference10. The refractive index of the medium was assumed to be constant and

equals 1.521 for PVA11 and 1.333 for water. Figure 6.3 shows the change of LSP as

a function of refractive index for two different aspect ratios of gold NRs.

The experiment data show fewer shifts than the calculation for both cases.

The shift is greater for longer NRs (∆L/d=4.3=53nm, ∆L/d=6.3=92nm). This deviation



141

might be attributed to the difference in the refractive index between the PVA film we

made and the one in the literature. We also infer that the process of stretching

probably lead to a less dense host medium for the gold NRs.

6.3.2. Polarization dependent color filters

As mentioned in the introduction, polymer-gold NRs nanocomposites have the

potential for the color filters showing an absorption spectrum that depends on the

polarization direction of incident light, which allows for the generation of multiple

colors in a single electro-optical element.

We used gold NRs of well-defined aspect ratio so that the color is predictable

and reproducible. Figure 6.4 shows the experimental set-up for the polarization

studies.

Figure 6.5 shows the polarizated spectra of the films containing gold NRs with

L/d= 2.8. The spectra demonstrate that for polarization parallel to the draw direction,

the transverse absorbance maximum completely disappears when the film is stretched

3 times, whereas the longitudinal resonance is still present. The intensity of

longitudinal resonance slightly decreases as draw ratio increases because of the

decrease in film thickness. Using perpendicular polarized light, the longitudinal

absorbance peak completely disappears when the film is stretched 4 times, whereas



142

the transverse resonance is still present. Since we obtained the complete

disappearance of longitudinal peak of transverse peak at the draw ratio above 4, we

use this ratio for the rest of the experiment.

Figure 6.4. The experimental set-up of the polarization spectroscopy studies. At a

polarization angle θ=0°, the electric field of the light is polarized parallel to the

stretch direction, whereas, at a polarization angle θ=90°, the electric field of the light

is polarized perpendicular to the stretch direction.

Figure 6.5. The polarization spectra of the gold NRs with L/d=2.8 dispersed in PVA

for various elongation numbers with respect to the original film. Indicated on the

spectral curves. a) θ=0°, b) θ=90°.

a) b)

450 600 800 1000

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4


Wavelength(nm)

2

3

4

450 500 550 600 650 700 750 8000.0

0.2

0.4

0.6

0.8

1.0


Wavelength(nm)

2

3

4



143

The alignment of gold NRs was also confirmed by confocal microscopy.

Figure 6.6 compares the alignment of gold NRs obtained by stretching. The random

distribution of single rods is observed in the unstretched film (a). When stretched, a

strong preferential orientation in the stretch direction is present.

Figure 6.6. CSLM micrographs of the PVA film containing gold rods with L/d = 7demonstrating alignment with stretching (a) the unstretched film and (b) stretched

film.

Figure 6.7 shows the absorption spectrum at various intermediate polarization

angles. As the angle changes from 0° to 90°, intensity of longitudinal plasmon starts

to decrease and that of transverse starts to increase. LSP appears at the longest

wavelengths when the angle is 0°and is blue-shifted as the polarization angle changes

to 90°. These results are in general accord with the calculation and the experimental

results of other groups2 . Changes from blue to red were observed by switching the

polarizer from the parallel to the perpendicular polarization direction, respectively.

a) b)

20 µm

S t

r e t c h i n g

d i r e c t i on



144

400 600 800 1000

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

90o

0o

90o

0o


Wavelength(nm)

Figure 6.7. UV-vis-NIR spectra of PVA/gold NRs nanocomposites for varying

polarization angles. L/d of gold NRs is 2.8.



direction parallel to the drawing direction and c) polarization direction

perpendicular to the drawing direction. Scale bar is 50µm.

We obtained transmittance spectra at different polarizer angle (Figure 6.9).

Based on the spectra, extinction ratio was calculated.

[dB])/T(Tlog10ratioExtinction //10 ⊥=

where T⊥ and T// are the transmittance perpendicular and parallel to the stretching



145

direction, respectively. Maximum extinction ratio was 18 dB at λ=LSP which is

considered to be fairly good, compared to those previously reported in the literature12.

The thickness of the film is 50µm and it has good flexibility.

420 600 800 1000

0

10

20

30

40

50

60

70

80

90

100

18dB @ 690nm

0o

90o

T r a n s m i t t a n c e

Wavelength(nm)

Figure 6.9. Transmittance spectra of PVA/gold NRs nanocomposites for varying polarization angles. L/d of gold NRs is 2.8.

To obtain films with vivid colors in the visible region, shorter NRs with

L/d=2.5 were mixed with larger nanosphere (d=30nm) and used to prepare the

nanocomposites film. The UV-Vis spectra of drawn nanocomposites in linearly

polarized light are shown in Figure 6.10. When the angle is 0°, we observed not

only the LSP but also the peak from the sphere which was not affected by the

stretching of the film. When the angle is 90°, broad peak at around 540nm was

observed which was attributed to the TSP of NRs and nanospheres. The LSP was



146

not completely gone in this case, since the aspect ratio is lower.

400 600 800 1000

0.00.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

90o

Unpolarlized

0o


Wavelength(nm)

Figure 6.10. UV-Vis-NIR spectra of PVA/gold NPs nanocomposites for varying

polarization angles. Gold NPs consist of NR of L/d= 2.5 and nanosphere of

diamerter=30nm.


gold NPs mixture and draw ratio 4. a) unpolarized, b) polarization direction parallel to

the drawing direction and c) polarization direction perpendicular to the drawing

direction. Scale bar is 50µm.

Figure 6.11 shows the color of the film under the different polarizer angles.

Polarization angle dependence of color is apparent.

When the aspect of NRs is sufficiently large, the LSP shifts to the near-IR

region. This means that the wavelength region displaying optical dichroism can be



147

shifted from the visible to the near-IR. This enables the fabrication of thin-film

optical filter that responds to the wavelengths in the near IR region. We probe this

possibility by preparing the film with longer aspect ratio of NRs.

400 600 800 1000 1200

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8


Wavelength(nm)

Figure 6.12. UV-Vis-NIR spectrum of PVA/gold NRs nanocomposites film. Theaspect ratio of NRs is 6.3.

Figure 6.12 shows the UV-vis-NIR spectrum of the nanocomposites film

prepared by using the NRs of aspect ratio 6.3. In the visible region, only the

transverse peak contributes to the color of the film. When the film is stretched, the

UV-vis-NIR spectra show the polarization dependence (Figure 6.13). The spectra

were split to 2 wavelength regions because of the use of different polarizers. When

the angle is 0, transverse peak is completely disappears as expected. The LSP was

appeared at longer wavelength. When the angle is 90°, the longitudinal peak is



148

suppressed and only transverse peak was observed.

Figure 6.13. The polarization spectra of the gold NRs with L/d=6.3 dispersed in PVA.

a) Visible region and b) near IR region.

Figure 6.14 shows the color of the films. Since there is no characteristic peak

in the visible region when the angle is 0, the film looks transparent. (Figure 6.14,b).



direction parallel to the drawing direction and c) polarization direction

perpendicular to the drawing direction. Scale bar is 50µm.

a) b)

350 400 450 500 550 600 650 700 750 800

0.0

0.1

0.2

0.3

0.4


Wavelength(nm)

parallel perpendicular

800 850 900 950 1000 1050 1100 1150 1200 1250 1300 1350

0.0

0.1

0.2

0.3

0.4


Wavelength(nm)

parallel

perpendicular



149

6.4. Conclusions

We prepared gold NRs /PVA nanocomposites film and obtained the alignment

of gold NRs in the PVA film d by simple stretching film method. The possibility of

using this film as polarization dependant color filters was demonstrated. The

polarization–dependent optical response of the NRs was measured and we confirmed

that good agreement was found with the theoretical prediction and previous studies.

By using higher aspect ratio of NRs, we observed strong optical dichroism in the near

– IR region which can be utilized to make optical device in near-IR region. We

believe that gold NRs provide an alternative to conventional, organic polarizers

because of their tunable peak positions and higher photo stability.



150

References

1 Van der Zande, B.M.I ; Koper, Ger J. M.; Lekkerkerker, Henk N. W. “Alignment of

Rod-Shaped Gold Particles by Electric Fields” J. Phys. Chem. B 1999, 103, 5754.

2 Carotenuto, G.; Her, Y,; Matijevic, E. “Preparation and Characterization of

Nanocomposite Thin Films for Optical Devices” Industrial & Engineering

Chemistry Research 1996, 35, 2929.

3 Al-Rawashdeh, N.A.F. ;Foss, Jr., C.A. “UV/visible and infrared spectra of

polyethylene/nanoscopic gold rod composite films: Effects of gold particle size,

shape and orientation” Nanostructured materials 1997, 9, 383.

4 Al-Rawashdeh, N.A.F.; Sandrock,M.L; Seugling, C.J,;Foss, Jr., C.A. “Visible

Region Polarization Spectroscopic Studies of Template-Synthesized Gold

Nanoparticles Oriented in Polyethylene” J. Phys. Chem. B 1998, 102, 361.

5 Dirix, Y.;Darribere, C. Heffels, W.; Bastiaansen, C.; Caseri, W.; Smith, P.,

“Optically anisotropic polyethylene-gold nanocomposites” Applied Optics

1999, 38, 6581.

6 Caseri,W. “Nanocomposites of polymers and metals or semiconductors: Historical

background and optical properties” Macromol. Rapid Commun. 2000,21, 705.

7 van der Zande,B.M.I.; Pages,L.; Hikmet, R.A.M.;van Blaaderen, A. “Optical

Properties of Aligned Rod-Shaped Gold Particles Dispersed in Poly(vinyl alcohol)

Films” J. Phys. Chem. B 1999, 103, 5761.

8 Murphy, C.J.; Orendorff C.L. “Alignment of gold nanorods in polymer composites

and on polymer surfaces” Adv. Mater. 2005, 17 ,2173.

9 Pérez-Juste, J.; Rodríguez-González, B.; Mulvaney, P.; Liz-Marzán, L. M. “Optical

Control and Patterning of Gold-Nanorod-Poly(vinyl alcohol) Nanocomposite Films”

Adv. Func. Mater. 2005,15, 1065.

10 Johnson P. B.; Christy, R. W. “Optical Constants of the Noble Metals” Phys. Rev.

B, 1972, 6, 4370.



151

11Cheremisinoff, N. “Handbook of Engineering Polymeric Materials” Marcel Dekker

New York 1997

12 Matsuda, S.;Yasuda,Y.;Ando,S. “Fabrication of Polyimide-Blend Thin Films

Containing Uniformly Oriented Silver Nanorods and Their Use as Flexible , Linear

Polarizers” Adv.Mater .2005, 17 ,2221.



152

7. Chapter VII

Pattern Formation of Gold Nanorods by Drying-Mediated Self

Assembly

Abstract

In this chapter, we investigate the microscopic patterns formed by evaporation

of gold NR suspensions in water. The drying of thin films of gold NR solution creates

structured patterns such as ribbon, ring and cellular network, depending on the

concentration (surface coverage) and evaporation rate. We probe the mechanism to

form such a pattern and analyze the structure to provide fundamental knowledge to

design statistically patterned arrays of nanoparticles and fabricate spontaneously

organized nanoscale devices.



153

7.1. Introduction

One of the growing interest in nanoparticles (NPs) research is to use NPs as

building blocks for constructing optoelectronic, electronic and magnetic devices1. To

achieve this goal, individual NP has to be organized into a specific structure. Among

various methods of making nanostructures, self assembly is considered to be one of

the few practical strategies because it is simple, versatile, and cost-efficient2. And in

some occasions, one can take advantage of spontaneous development of patterns of NPs

as a convinient path for sophisticated products.

In the self assembly process, it is important to understand the various

interactions between NPs, substrates, and solvents, which lead to the patterns formed.

There are several competing interactions/forces between particles that control the self

assmebly behavior, such as Brownian motion, electrostatic attraction and repulsion, van

der Waals attraction, steric repulsion, and capillary force, etc3. Since most of the self

assembly is accomplished by evaporating thin film of solution on a substrate, drying

kinetics, substrate roughness, solvent wetting/dewetting, hydrodynamic effects and self–

diffusion of the NPs on a substrate also play important roles in governing the

morphologies of self aseembly leading to unusual nonequilibrium structures4.



154

Evaporation of solvent affects particle/particle interaction, since the relatively

weak attraction between NPs, which are efficiently screened in dilute solution,

becomes noticeable as the solvent evaporates, initiating assembly of slowly evolving

structures5.

It is observed that for sufficiently monodisperse systems, hard colloidal spheres

self assemble into a wide range of highly orderd phases due to entropic effects6. The

driving force of this self assemly is repulsive interactions between colloidal spheres.

At high volume fractions, the extra packing entropy due to local motions about

lattice sites compensates for the loss of configurational entropy arising from long-range

ordering. The factors affecting the mechanism are the concentration of the particles,

softness of the interparticle potential, the existence of slight interparticle attractions and

particle polydispersity.

Ohara et al. observed a solution of sufficiently polydisperse gold nanocrystals

self assembled showing size-segregation during solvent evaporation, suggesting that the

size-dependent attraction between the nanocrystals is sufficient to induce size

segregation7 since entropic effect is suppressed due to the polydispersity.

Ge et al. suggested that self assembly is the result of aggreration from a phase

separation between a dense nanocrystal liquid and dilute nanocrystal vapour (phase



155

condensation model) that is triggered by the increase of interaction between NPs when a

good solvent is replaced by a poor solvent (air)8. They showed a variety of structured

aggregate spatial patterns as a function of surface coverage from the drying of thin

nanocrystal solution film.

Rabani et al. further extended phase condensation model to present a model of

NP self assembly which included the dynamics of the evaporationg solvent5. The

simulations account for all observed spatial and temporal petterns including network

structure. They categorized two distinct mechanisms of ordering, corresponding to the

homogeneous and heterogeneous limits of evaporation dynamics. In the case of

homogeneous evaporation, evaporation is spatially homogeneous while the boundaries

of nanoparticle domains remain fluxional throughout the growth dynamics.

Depending upon coverage and the evaporation condition, different patterns were

observed (Figure 7.1).



156

Figure 7.1. Self-assembled morphologies resulting from homogeneous evaporation

and wetting of nanoparticle domains. The upper panels show the experimental results.

The lower panels show results of model simulations for coverages of 5% (a), 30% (b),

40% (c) and 60% (d). The upper panels show corresponding experimentally observed

morphologies for CdSe nanocrystals at similar coverages. From reference 5.

In the case of heterogeneous evaporation (evaporation is instead strongly

heterogeneous in space), the dynamics of self assembly can be dramatically different.

Phase change occurs not by long-wavelength fluctuations, but by nucleation and growth

of vapor bubbles leading to network-like morphologies (Figure 7.2). These

heterogeneous structures are highly versatile and used both in nature and industry. The

connectivity of the phases is one of the key factors controlling the transport functions

and the mechanical properties of the phase separated structures9.



157

Figure 7.2. Self-assembled morphologies resulting from inhomogeneous evaporation in

simulations (left) and in experiments (right). From reference 5.

Ohara et al. further observed annular ring-like arrays when dilute solutions of

nearly monodisperse metal NPs were evaporated on a solid substrate10. Each annular

ring is argued to arise from the pinning of the rim of an opening hole.

Figure 7.3. a) A ring-like array formed from the dilute solution of nanocrystals. b) Side-

view schematic showing a scenario for evaporation of solvent from nanoparticle

solution on wetted substrate (top), and its thinning to thickness thole < te where thole is the

thickness of the film and te is the critical thickness at which point a hole nucleates

(middle). The hole then opens, collecting particles in its growing perimeter (bottom),

which becomes an annular ring upon pinning of its contact line. From reference 10.

a) b)



158

Capillary-forces-induced self assembly is another way to make a large scale

configuration11. Capillary forces are interactions between particles mediated by fluid

interfaces12. As the thickness of the water layer becomes equal to the particle diameter,

a nucleus of 2D crystal suddenly forms. The particles surrounding the nucleus begin

to move towards the ordered phase to join the ordered array. The quality of assembly

is determined by the factors such as the rate of water evaporation, the ionic strength, the

surfactant concentration and the particle size and shape distributions.

For the self assembly of anisotropic shape of NPs such as nanorod (NR),

nanowire, and nanodisk, the situation becomes even more complex, because they can

form liquid crystalline phases. Although a great deal of work has been published on

the self assebmly of spherical NPs, to date the assembly behavior of anisotropic

particles is largely unexplored.

This chapter will focus primarily on observations of self assembly of gold NRs

and will probe the mechanism for the formation of different patterns.

7.2. Experimental

We obtained gold NRs of spherocylinderical shape with controlled aspect ratio

and with a little byproduct (20% by number fraction) by using seed-mediated method



159

with binary surfactant solution according to the procedures described in chapter 2. We

successfully separated NRs from spherical NPs employing a single round of

centrifugation (described in chapter 4). And excess surfactant was removed by

additional centrifugation. Thus the final sample contains NRs with low polydispersity

(less than 15%) .

For each aspect ratio of gold NRs, the concentration is varied by adding distilled

water to the mother solution. The concentration of gold NRs to form the monolayer on

the grid by deposition of 1µl solution is roughly 1×10-13 mole/liter (for the NRs with

aspect ratio of 6). We fix the concentration of the mother solution at 2×10-12 mole/liter

so that the observation is not limited by the thick multi layer of the self assembly.

All TEM samples were prepared by depositing 1µl of the solution of specific

concentration on TEM grids. Since we used carbon coated copper grid, the deposited

solution did not spread and remained stable, hence, the complete evaporation took

almost 1hr (condition I). Evaporation was slowed down by placing platform of the

grid in the water bath at room temperature and covering it with a glass. The water

provided a water vapor atmosphere to allow for slow drying (6-12 h) of the solutions on

the grid (condition II). We set the initial film thickness by removing the excess

solution with filter paper 3 minutes after deposition. In this case, the complete



160

evaporation took less than 10 minutes. The nature of the self assembly of NPs was

examined by TEM (JEOL100 at 100 KV).

7.3. Results and discussion

7.3.1. Interparticle forces among NRs.

Before we proceed to discuss the self assembly, we need to understand the

interactions among gold NRs in solution. The interaction between colloidal particles

consists of both electrostatic repulsive forces due to the overlap of electrical double

layers and of attractive van der Waals forces. As-made gold NR in water is stable up

to considerably long time (several months) because it is covered by the bilayer of

surfactant. The governing interparticle force in this case is electrostatic repulsion due

to the electrical double layer and steric hindrance. According to our experimental

observation, upon repeated centrifugation (usually 3 times), gold NRs aggregate and it

is impossible to redisperse them. In this case, the governing interparticle force can be

considered to be van der Waals attraction. Therefore, the concentration of surfactant

remaining in the solution is critical because the stability of colloidal suspensions

depends on the balance between these two forces. We prepared the mother solution for

self assembly by centrifuging as-made solution twice. The final concentration of



161

surfactant is about 1×10-3 M. This sample was stable for several months. Thus the

colloid-colloid interactions can be approximated by the sum of hard-core, electrostatic

(soft) repulsions and van der Waals attraction. And the dominant interaction is

determined by the distance between the particles.

7.3.2. Heterogeneity and polydispersity of the sample

As-made solution is essentially a mixture of rods and spherical particles. We

observed two different phenomena when as-made solution is evaporated on a TEM grid.

When the evaporation is relatively slow (condition II), we observe shape-selective

assembly where NRs and spherical particles assemble separately showing micro phase

separation on the TEM grid.

Figure 7.4. TEM image of self assembly of as-made solution. (a) Fast evaporation

(condition I) and (b) slow evaporation (condition II). The coverage of the particles in

both images is about 20%.

Figure 7.4 (b) shows the shape dependent phase separation. This is driven by

a) b)



162

depletion interactions which can be explained in terms of purely repulsive interactions,

between particles with different shape. This phenomenon is extensively studied

theoretically and has also been confirmed by simulations. The corresponding “depletion

potential” is given by13

3

r B ) L

h1(

D

R

D

LT k

3

2)h(W −−= φ

where k BT is the thermal energy, φr is the volume fraction of the rod, R is the radius of

the sphere. L is the length of the rod, D is the diameter of the rod, and h is the distance

between the surfaces of the two particles. We observe this separation for NRs of

relatively high aspect ratio, since high L/D is one of the favorable conditions.

When the evaporation is rather fast, we do not observe this micro phase

separation. We think that in this case, the phase separation could not be achieved

because the particles lose their mobility since the evaporation is completed in a short

time.

Figure 7.5 is another example showing micro phase separation of 3 different

shapes of NPs (nanorods, nanospheres and nanocubes).



163

Figure 7.5. TEM image of self assembly of as-made solution showing micro phase

separation of 3 different shapes of nanoparticles.

The polydispersity within the same species is also an important parameter

affecting the quality of self assembly. NPs with higher polydispersity do not show any

regular assembly under any circumstance because polydispersity usually prevents long

range ordering

14

. Even on the same TEM grid, local polydispersity does determine the

character of self assembly.

There are several reports regarding self assembly of gold NRs15, focused on the

local self assembly in a small domain. It has been proposed that the concentration and

pH of the solution determine the properties of self assembly15,c. Using TEM, we can

observe most of the characteristic patterns both from homogeneous evaporation and

heterogeneous evaporation at every level of coverage ratio within a single sample

prepared on a TEM grid.



164

Scheme 7.1. The schematic showing the preparation of TEM samples.

We prepared the sample for TEM observation by deposition of 1µl gold NR

solution on a TEM grid. The radius of the drop is smaller (1.4 mm) than that of the grid

(3mm). This condition enables us to observe the coffee ring phenomena confined on a

TEM grid. The colloidal particles migrate to the edge where the contact line of the

drop gets pinned, and the evaporating fluid at the edge is replenished by a strong

capillary driven flow that carries nearly all the solute to the edge. The solute becomes

concentrated into a fraction of the perimeter during evaporation (We will discuss this

phenomena in detail in chapter 8). As a result, on the same TEM grid, the

concentration of the solute (gold NPs) increases towards the edge of the grid.

Therefore we can observe the concentration dependant pattern formation on the same

TEM grid. Also evaporation rate is higher towards the edge providing different

evaporation rates on the same grid. Because the local environment such as roughness

of the surface and thickness of solution film formed on the grid could not be controlled

Gold NR solution

Glass slide

TEM grid



165

even on the same TEM grid, we observe the characteristic patterns both from

homogeneous and heterogeneous evaporation.

We discuss the patterns formed by gold NRs depending on the local coverage of

gold NRs rather than the concentration since it is impossible to determine the local

concentration of the solute. Also, the self assembly observed on a TEM grid is mostly

monolayer, therefore it can be considered as two dimensional. The following results

were obtained by slow evaporation (condition II).

7.3.3. Patterns formed by homogeneous evaporation

Figure 7.6 shows the patterns taken from the center area of a single TEM grid.

The aspect ratio of NRs is 5. At very low coverage (less than 5%) ribbon-like

aggregates of nanoparticles dominate the self assembly (Figure 7.6 a). As the

coverage increases (Figure 7.6 b), disk-like aggregates of NPs dominate the self-

assembly. At higher coverage, NR domains become multilayer anisotropic shape

(Figure 7.6 c) and finally become connected cluster, when the particle coverage is

higher, around 20-30% (Figure 7.6 d).

Our results are qualitatively consistent with the dynamics of NP assembly in the

limit of spatially homogeneous evaporation where the interparticle attractive force,



167

Figure 7.7. TEM images of self assembly of gold NRs showing transition from ribbon-

like structure to disk-like smectic domain. a) The individual domain of ribbon-like

structure, b) small domain of nematic-like structure, c) domains of smectic-like

structure and d) an enlarged image of c showing an individual domain of smectic-like

structure.

We also found that as the size of domain (aggregate) increases, the assembly

shows the phase transition. We mentioned previously that at low coverage, side by

side assembly of gold NRs is dominant. Figure 7.7 a) shows the individual domain of

ribbon-like structure. As the size of domain (also the number of particles in the

domain) increases, multi layer of ribbon structure is observed as well as monolayer and

a) b)

c) d)



168

ribbons transform into small domain of nematic like structure shown in Figure 7.7 b)

and finally into smectic-like structure as shown in Figure 7.7 c).

The ribbon-like structure is simply due to the anisotropy of NRs. Bates et al.

studied two dimensional hard rod fluids consisting of spherocylinders confined to lie in

a plane by means of Monte Carlo simulations 16. They predicted that due to the

anisotropy of NRs, chains of particles aligning side-by-side form ribbon-like structure.

For long rods, a 2D nematic phase and smectic phases are observed at higher density.

Our experimental observation is in good agreement with their simulation showing that

the transition to nematic like structure and smectic like structure occurs in the domain of

aggregation.

In the same region, we often observed a considerably large area covered by

monolaryer of NPs, sometimes forming interesting features (Figure 7.8).



169

Figure 7.8. a) TEM images of self assembly of gold NRs showing a large area covered

by monolayer of gold NRs. b) One of the interesting features formed by the

monolayer of gold NRs. The “eye area” is the microphase-separated nanospheres

surrounded by NRs.

7.3.4. Patterns formed by heterogeneous evaporation

One of the interesting patterns we observed from evaporation of gold NR

suspension on the TEM grid were the ring pattern. This sturcture was observed when

a)

b)



170

the coverage is 10 to 20% and at the intermediate area between the center and the edge

of the drop.

Figure 7.9. Ring-like array observed when the coverage is 10 to 20%. a) aspect ratio 4,

b) aspect ratio 6.

The ring patterns formed by various aspect ratio of gold NRs (aspect ratio 4 to 7).

The shape of the ring is close to a perfect circle and the diameter of the ring ranged from

200nm to 1µm and not related to the aspect ratio of gold NRs. The neighboring rings

are alingned forming a line which could be a part of the circumference. Their size is

almost same and the distance between each ring is also very regular.

a)

b)



171

The ring patterns result from the pinning of the contact lines of nucleating

holes in the thin liquid film. The hole opens, pushing most particles out along its

advancing rim,which becomes a ring. Depending on the local concentration of particles,

the width of the outer rim varies. The reason for the set of ring forming a line is

related to the concentration gradient formed in the evaporating drop. After the

contact line gets pinned, colloidal particles migrate to edge. This causes dynamic

concentration gradient of solutes along the radius of the drop. As the water evaporates,

the film gets thinner. Below its critical thickness (te), it becomes unstable resulting in

the nucleation and growth of holes. Since the condition such as the concentration of

solute as well and the thickness of the solution film is radially similar, holes are formed

along a macroscopic ring which can be recognized as a line in nanoscale observation.

We measured the volume of the particle within a ring-like array and found that it

decreases toward outer rim which indicates that lighter particle is pushed farther than

heavier particle (Figure 7.10). This result supports the mechnism that the ring

originates from the hole instability.



172

0

2000

4000

6000

8000

10000

12000

100 120 140 160 180 200 220

distance from the center(nm)

v o l u m e o f t h e p a r t i c l e s ( n m

3 )

Figure 7.10. The volume of the particle within a ring-like structure decreases towardouter rim which indicates that lighter particle is pushed further than heavier particle.

We also observed circular domains which do not have the hole in the center

(Figure 7.11). This structure is similar to the structure formed by percolation of the

neighboring holes17. When the rims of holes are not pinned, neighboring holes

percolate to form a “drop” at their interstices, pushing particles together (middle),

resulting in close-packed circular aggregation.

Figure 7.11. The circular domain formed between the ring-like array structures.



173

Figure 7.12. TEM images of self assembly of gold NRs showing the formation of the

ring-like structure in different patterns.

Figure 7.12 shows the interesting ring array inside the uniform monolayer of self

assembly. Figure 7.13 shows another interesting ring array having the monolayer of

self assembly as a background. This is the superposition of the two opposite sides (top

and bottom) of the TEM grid (both sides of the TEM grid are visualized without

distinguishing one from the other). Since the solution smears out to the bottom side of

the grid during the deposition, similar phenomena occur on both sides. One side

shows the uniformly formed monolayer of NRs while the other side shows the ring-



174

array structure. We assume that the different local environment is responsible for

these different structures.

Figure 7.13. TEM image of ring array having the monolayer of self assembly as a

background.

As the coverage increases, ring arrays become cellular networks. Figure 7.14

shows the coexistance of ring-like array and the intermediate state between ring-like

array and network structure.

Figure 7.14. TEM image showing the transition from ring-like array to cellular

network structure.



175

The celluar network structures are often observed when the coverage reaches

about 30 to 50%. Figure 7.14 shows the transition from ring-like array to cellular

network structure.

Figure 7.15. The cellular network structures formed by gold NRs with different aspect

ratios: a) aspect ratio of 4 and b) aspect ratio of 6.

Figure 7.15 shows the network structure formed with different aspect ratios. The

edge of inside cell is better defined for the shorter NRs. The size of the hole ranges

a)

b)



176

from 100 to 700nm in diameter and is not related to the aspect ratio of NRs. We believe

that the difference in the size of the hole originates from the locally different

environment such as local film thinkness fluctuation, surface roughness and small

temperature differences.

Figure 7.16. Cellular network structure formed in a large area.

The network structure is very useful because it gives the connectivity of the

phase at the expense of small amount of materials. We observed that gold NRs can

form this network structure over a considerably large area (Figure 7.18) showing the

potential applications as in fabricating electronic devices.

7.3.5. Self assembly formed at the edge of the evaporating drop

Figure 7.17 shows a large area of the assembled NRs on the dried edge of the

drop. The outer edge consists of monolayer of NRs in which NRs assembly does not



177

show specific ordered phase. The NRs at the interface of air-solution seem to lie

parallel to the interface. Moving inside from the edge, the second region shows more

ordered structure where longer chains of NRs are densely packed forming multilayer.

Further away from this region, domains of smectic-like assembly of NRs are formed

less densely.

The anisotropy of NR is the reason for forming this assembly. As the particles

transfer towards the edge by the convective flow, the local concentration reaches high

enough for phase transition from isotropic to nematic and smectic. The domains of

smectic LC phase joins to the edge and the strong lateral capillary force align them into

dense ordered structure

Figure 7.17. The self assembly of gold NRs formed at the edge of the evaporating drop.



178

Figure 7.18. Different patterns of self assembly formed by evaporating gold NR

solution.

The various patterns are mapped as a function of distance from the center of the

drop and the coverage (Figure 7.18). We observe the patterns similar to those from

homogeneous evaporation in the center of the drop and various patterns similar to those

from heterogeneous evaporation towards the edge of the drop. These observations show

that the concentration of NRs, the evaporation rate, the mobility of the particles and the

thickness of the solution film are the key factors to determine the properties of self

assembly.



179

7.3.6. Effect of molecular linkage on the self assembly of gold NRs

In nanomaterial research, there is interest in the bulk fabrication of

supramolecular nanostructures where individual NPs are arranged in a controlled

fashion. Optimally, the NPs can be naturally assembled into the final structure through

particle-particle interaction as we aforementioned. Supramolecular assembly of NPs

can be achieved by controlled association through complementary recognition of NP

surface functionality. In the case of gold NRs, end-to-end assembly has been

demonstrated using molecular linkage such as α-ω-alkanedithiols18, cysteines19 and

biotin-streptavidine connectors20. The dense adsorption of surfactant on the side of

NR is responsible for this end- to-end assembly.

We also achieved longitudinal assembly through cooperative intermolecular

hydrogen bonding by using molecules such as azo dye rotaxane. By changing the

concentration of azo dye rotaxane, we were able to control the number of particles

linked together (Figure 7.19).



180

Figure 7.19. TEM images of gold NRs in the presence of azo dye rotaxane showing

end-to-end linkage mediated by azo dye rotaxane. The aspect ratio of the gold NRs is

4.5. The concentrations of dye rotaxane increases from a to d. a) 2.0×10-6M, b)

4.08×10-6M, c) 4.9×10-5M and d) 7.35×10-5M.

7.4. Conclusions

A variety of patterns were formed by evaporating solutions of NRs deposited on

TEM grids. The patterns can be explained by drying-mediated self asssembly where the

coverage ratio and evaporation condition are among the dominant factors that determine

the characteristics of the pattern. Depending on the location in the drop, the

concentration of the gold NR and the evaporation rate are different, thus self assembly

shows the chasracteristic patterns from both homogeneous evaporation and

heterogeneous evaporation. The quantitative analysis of the specific mechanism of

self assembly needs to be studeid further to design statistically patterned arrays of

Gold Nanorod

Azo dye rotaxane

(a) (b) (c) (d)



181

nanoparticles and to fabricate spontaneously organized nanoscale device. We believe

that by understanding the mechanism of the self assembly and by achieving the

controlled evaporation environment, we will be able to use the self assembly as a

vasatile nonlithographic technique having potential for large scale fabrication of the

next generation of inexpensive electronic devices.



182

References

1

El-Sayed, M.A. “Some Interesting Properties of Metals Confined in Time and Nanometer Space of Different Shapes” Acc.Chem. Res. 2001, 34, 257.

2 Whitesides, G. M.; Grzybowski, B. “Self-Assembly at All Scales” Science,2002, 295,

2418.

3 Hiemenz, P.C.; Rajagopalan, R. “Principles of Colloid and Surface Chemistry” 3rd ed.,

1997, Marcel Dekker, Chapter 10 and 11.

4 Tang,J.; Ge, G.; Brus, L.E. “Gas-Liquid-Solid Phase Transition Model for Two-

Dimensional Nanocrystal Self-Assembly on Graphite” J. Phys. Chem. B 2002, 106,

5653.

5 Rabani, E.; Reichman D.R.; Geissler, P.L.; Brus L.E. “Drying-mediated self-assembly

of nanoparticles” Nature 2003, 426 , 271.

6 a) Pusey, P. N.; van Megan,W. “Phase behavior of concentrated suspensions of nearly

hard colloidal spheres” Nature 1986, 320, 340 b) van Megan,W; Underwood S. M.

“Glass transition in colloidal hard spheres: Mode-coupling theory analysis” Phys. Rev.

Lett. 1993, 70, 2766 c) Lin,K.; Crocker,J.C., Prasad,V.; Schofield,A.; Weitz, D. A.;

Lubensky, T. C.; Yodh, A. G. “Entropically Driven Colloidal Crystallization on

Patterned Surfaces” Phys. Rev. Lett. 2000 85,1770.

7 Ohara,P.C.; Leff, D.V.; Heath,J.R.; Gelbart,W.M. “Crystallization of Opals from

Polydisperse Nanoparticles” Phys. Rev. Lett . 1995,75, 3466.

8 a) Ge, G.; Brus, L. E. “Evidence for spinodal phase in two-dimensional nanocrystal

self-assembly” J. Phys.Chem. B 2000,104, 9573 b) Tang,J.; Ge, G.;Brus,L.E. “Gas-

Liquid-Solid Phase Transition Model for Two-Dimensional Nanocrystal Self-

Assembly on Graphite” J. Phys. Chem. B 2002, 106, 5653.

9 Iwashita,Y.; Tanaka, H. “Self-organization in phase separation of a lyotropic liquid

crystal into cellular, network and droplet morphologies” Nature materials 2006, 5,

147.



183

10 Ohara, P.C.; Heath,J.R.; Gelbart,W.M. “Self-Assembly of Submicrometer Rings of

Particles from Solutions of Nanoparticles” Angew. Chem. Int. Ed. Engl. 1991, 36 ,1077.

11 Nikoobakht, B; El-Sayed, M. A. “Self Assembly of Gold Nanorods” J.Phys.Chem.

B . 2000, 104, 8635.

12 Kralchevsky, P.A.;Denkov,N.D. “Capillary forces and structuring in layers of colloid

particles” Current Opinion in Colloid and Interface Science 2001, 6 , 383.

13 Koenderink,G.H.; Vliegenthart, G. A.; Kluijtmans, S. G. J. M.; van Blaaderen,

A.;Philipse, A. P.; Lekkerkerker, H. N. W. “Depletion-Induced Crystallization in

Colloidal Rod-Sphere Mixtures” Langmuir 1999, 15, 4693.

14 Gabriel1, J.P.; Davidson, P. “Mineral Liquid Crystals from Self-Assembly of

Anisotropic Nanosystems” Top. Curr. Chem. 2003, 226 , 119.

15 a) Jana,N.R. “Shape Effect in Nanoparticle Self-Assembly” Angew. Chem. Int. Ed .

2004, 43, 1536 b) Sau,T.K.; Murphy,C.J. “Self-Assembly Patterns Formed uponSolvent Evaporation of Aqueous Cetyltrimethylammonium Bromide-Coated Gold

Nanoparticles of Various Shapes” Langmuir 2005, 21, 2923 c) Orendorff,C.J.;

Hankins, P.L.; Murphy,C.J. “pH-Triggered Assembly of Gold Nanorods” Langmuir

2005, 21, 2022.

16 Bates, M.A.; Frenkel,D. “Phase behavior of two-dimensional hard rod fluids” J.

Chem. Phys. 2000, 112, 10034.

17 Ohara,P.C.; Gelbart,W.M. “Interplay between Hole Instability and Nanoparticle

Array Formation in Ultrathin Liquid Films” Langmuir 1998, 14, 3418.

18 Joseph,S.T.S.; Ipe,B.I.; Pramod, P.;K. George Thomas,K.G. “Gold Nanorods to

Nanochains: Mechanistic Investigations on Their Longitudinal Assembly Using αϖ-

Alkanedithiols and Interplasmon Coupling” J. Phys. Chem. B 2006, 110, 150.

19 Sudeep, P. K.; Joseph, S. T. S.; Thomas, K. G. “Selective Detection of Cysteine and

Glutathione Using Gold Nanorods” J. Am. Chem. Soc.2005, 127 , 6516.



184

20 Caswell, K. K.; Wilson, J. N.; Bunz, U. H. F.; Murphy, C. J. “Preferential End-to-

End Assembly of Gold Nanorods by Biotin-Streptavidin Connectors” J. Am. Chem.

Soc. 2003, 125, 13914.



185

8. Chapter VIII

Lyotropic Liquid Crystalline Phases of Colloidal Gold Nanorods

Observed in Evaporating Drop

Abstract

We accomplished the experimental observations of lyotropic liquid crystalline

(LC) phases of colloidal gold nanorods (NRs) by achieving the phase transition

volume fraction through convective flow resulting from solvent evaporation.

Characteristics of LC were confirmed by birefringence and fluidity of the LC domain

and complemented by the self assembly observed by TEM. LC phase was observed

for nanorods with aspect ratios ranging from 4 to 14. By changing the aspect ratio,

concentration , and evaporation rate, we observed various patterns formed by the

self assembly of gold NRs. When the dilute solution is evaporated fast, the contact

line experiences a series of pinning-depinning events, and the gold NRs accumulate

where the contact line is pinned resulting in the formation of multiple rings.



186

8.1. Introduction

Liquid crystalline (LC) phases exhibit properties of liquid and solid crystals.

The various LC phases can be characterized by the type of ordering that is present.

The nematic phase is one of the most common LC phases showing long-range

orientational order and no positional order. In the smectic phase, the molecules are

arranged in layers showing positional order in addition to orientational order. Liquid

crystals can also be classified into thermotropic and lyotropic LCs. Thermotropic

LCs go through a phase transition into the LC phase as temperature is changed,

whereas lyotropic LCs exhibit phase transitions as a function of concentration.

Scheme 8.1 schematically shows the phase transition of lyotropic liquid crystals due

to the change in the concentration.

Scheme 8.1. Various liquid crystalline phases as a function of concentration of the

solute.

Isotropic Nematic Smectic Solid

Concentration increases



189

from the observation of orientational and positional order in their assembly on the

TEM grid 12. This observation implies the formation of a liquid crystalline phase

prior to complete evaporation of the solvent.

However, the most challenging problem in studying lyotropic LC phase

formed by anisotropic NPs is to attain the critical concentration so as to obtain the

critical concentration necessary to observe the phase transition predicted by Onsager .

In the case of highly anisotropic (large L/d ratio) particles, concentrations of few

percent by weight are usually necessary in order to observe such a transition. With

particles of much lower aspect ratio, such as nanorods synthesized so far, much higher

concentrations are required.

Meanwhile, the LC phase of suspensions of CdSe semi-conductor nanorods

was explicitly demonstrated 13. The suspension of CdSe NRs of aspect ratios tunable

up to 15 shows schlieren textures in polarized light microscopy and displays

spontaneous isotropic/nematic phase separation in a given range of concentration.

LC phase of gold NRs may provide a promising feature since it combines the

typical properties of LC with the electro-optical response of colloidal dispersions14.

There has been an attempt to show LC phases of gold NRs15. However, LC

formation was not explicitly demonstrated.



190

8.1.2. Contact line deposits in an evaporating drop

When a drop of colloidal dispersion dries, the particles deposit in a ring at the

periphery of the drop. The colloidal particles migrate to edge where the contact line

of the drop gets pinned, and the evaporating fluid at the edge is replenished by a

strong capillary driven flow that carries nearly all the solute to the edge. The solute

becomes concentrated into a fraction of the perimeter during evaporation (Figure 8.1).

Figure 8.1. The factors leading to outward flow in a drop of fixed radius R slowly

drying on a solid surface. The evaporative flux J (r ) reduces the height h(r ) at every

point r . As its contact line is pinned, the radius of the drop cannot shrink. To prevent

the shrinkage, liquid must flow outwards v(r). From the reference 16.

Deegan et al.16 showed that these ring-like deposits can be observed whenever

the surface is partially wet by the fluid irrespective of the chemical composition of the

substrate. And they can be found with solutes ranging in size from the molecular to

the colloidal. By treating the droplet as a symmetric lens of fluid about the substrate

and making an analogy to electrostatics, the evaporative mass flux, J s, is given by

λ −−≈ )r R( )t ,r ( J s



191

where λ = ( π -2θ c )/(2π -θ c ), R is the droplet radius, r is the radial coordinate, and θ c is

the contact angle.

Figure 8.2 shows a dried coffee stain16 and a dried gold NRs solution. Since

gold NRs solution leaves a ring shaped deposit on the substrate gold NRs, we

rationalize that the outward flow carries gold NRs to the edge and the concentration of

NRs at the edge becomes high enough to undergo the phase transition from isotropic

to nematic liquid crystal.

Figure 8.2. Images of a) a 2-cm-diameter drop of coffee from reference 16 and b) a 4-

mm-diameter drop of colloidal gold NRs.

In this chapter, we accomplished the experimental observations of lyotropic

liquid crystalline phases of colloidal gold NRs by achieving the phase transition

volume fraction through the convective flow resulting from pinning of the

evaporating drop.

a) b)



192

8.2. Experimental

When NR suspension is treated as hard-rods17 , the critical concentration for

I-N phase transition is predicted to be 30 wt % for an aspect ratio of 10. Higher

aspect ratio of gold NRs are usually produced in a low yield with large size

distribution and byproduct (nanospheres and nanocubes). And it is relatively difficult

to attain the solution concentration higher than the critical concentration. Thus, LC

phase of gold NRs was not explicitly demonstrated.

We obtained spherocylindrical shape of gold NRs with controlled aspect ratio

and with little byproduct (20%) by seeded growth procedure with binary surfactant

solution according to procedures described in chapter 2. We successfully separated

NRs from the nanocubes employing a single round of centrifugation (detailed in

chapter 4). Excess surfactant was removed by additional centrifugation. Thus the

final sample is almost pure NRs with low polydispersity (standard deviation of aspect

ratio is less than 15%) and the volume fraction, Φ was 5 × 10-5. The desired

volume fraction is obtained by adding distilled water to the sample.

A specific amount of gold NR solution was deposited on a clean glass slide.

The evaporating solution was observed using an optical microscope under crossed

polarizers (Leica DMRX ).



194

0

5

10

15

20

25

30

35

40

45

0 50 100 150 200

time (sec)

c o n t a c t a n g l e ( d e g r e e )

Figure 8.3. The contact angle of a drop of gold NR solution (1µl) on a glass substrate

as a function of time.

Figure 8.3 shows the contact angle of a drop of gold NR solution on a glass

substrate as a function of time during the drying process. The contact angle linearly

decreases as water evaporates. When the contact angle becomes less than 5°, the

contact line starts to recede and finally water evaporates completely. First stage

occupies a substantial fraction of the drying time (1st stage ~168 sec and 2nd stage ~12

sec). The duration of this time is inversely proportional to the volume of the drop.

During this stage, a flow generated from the evaporating drop causes the solutes to be

carried to the edge resulting “coffee ring”. When the size of drop is the same, the

width of the overall ring (band width) linearly increases as the solute concentration

increases indicating that most of the particles are carried to the edge forming the ring

(Figure 8.4).



195

0

50

100

150

200

250

0 0.2 0.4 0.6 0.8 1

normalize solute concentration

b a n d w i d t h ( m m

)

Figure 8.4. The band width is plotted as a function of normalized volume fraction of

the gold NRs in the case of slow evaporation.

8.3.2. LC phase formed in an evaporating drop.

A simple way to identify a LC phase is to observe its texture using polarized

light microscopy19. Since LC phases are birefringent, they interact with linearly

polarized light, and can be observed between crossed polarizers.

We observed the coffee ring formation using polarized microscope. When

the aspect ratio of gold NR is less than 4, the ring formed at the contact line but we

didn’t observe any birefringence from the evaporating solution. The birefringence

from the surfactant crystal appears after water evaporates completely (Figure 8.5).



196

Figure 8.5. Optical microscope images of a dried drop of gold NRs with the aspect

ratio of 3. a) Bright field image and b) polarized image at the edge. Birefringence

is from the surfactant crystal.

When the aspect ratio is larger than 4, as the contact line is pinned, it becomes

highly birefringent. Then birefringent droplets emerge from the interior and move

towards the edge and join the existing birefringent contact line forming a ring. This

bright band moves inward with droptlets joining at the same time. Meanwhile, the

contact line separates from the outer edge and recedes inwards. After the contact line

recedes, birefringence of the area behind the contact line becomes weaker due to the

loss of fluidity. The formation of birefringent ring is shown in Figure 8.6.

The time to complete the evaporation is about 11min for a drop that is 2µm

in diameter. The highly birefringent edge appears only after 3min. This period can

be considered to be “self-pinning”20. Gold NRs wedge between the substrate and

the contact line.

a) b)



197

Figure 8.6. In this series of images the center of the drop is in the bottom left corner.

The images of the drop of gold NRs solution during evaporation under optical

microscope with cross polarizers, as described in the text. The scale bar is 30µm.

Figure 8.7. The images of the drop of gold NRs solution during evaporation underoptical microscope with cross polarizers, as described in the text. a) Emerging liquid

crystal domain from the interior near the edge and moving toward the edge. b)

Magnified image showing the individual domain joins the existing structure. c) The

assembly of liquid crystal domains before complete drying. The scale bar is 200 µm

for a) and 20 µm for b) and c).

The droplets in Figure 8.7 c) are about 3-4 µm in diameter. From the

birefringence and fluidity of the droplet, it is apparent that the droplets are domains of

t=3 mins t=4.5 mins t= 6 mins

t=7.5 mins t=9 mins t=10.5 mins

a) b) c)



198

liquid crystals21. Also the individual domain kept sparkling until the drop dried

completely. The sparkling of the individual domains or droplets is due to the

fluctuations of the optic axis of the nematic phase driven by thermal energy.

Figure 8.8. TEM images of a dried drop of gold NR solution. a) The edge of the drop

showing smectic-like structure. b) and c) show the small size of nematic-like domain

away from the edge.

The drop of gold NRs was evaporated on a TEM grid and taken to the TEM

observation. Figure 8.8 shows the self assembly of gold NRs in a dried drop. The

outer ege consists of monolayer of NRs and the NRs lie parallel to the interface of air

and the solution. Moving towards inside, the thick layer of smectic-like structure of

band width of a few hundred micrometers is found (a). Also, the reminiscence of

the nematic domain is found as shown in (b) and (c) away from the edge. These self

assemblies confirmed that as the concentration of gold NRs increases, NRs form a

small domain of nematic phase as they are draged towards the edge and go through

nematic-smectic transition as the domains join the existing structure.

a) b) c)



200

8.3.3. Two dimensional phase transition observed in self assembly on a TEM

To examine NR assembly systematically by TEM, a drop of 4 µl gold NRs

solution was deposited on a carbon coated copper TEM grids and the solution was

slowly evaporated in a closed environment. Evaporation rate is further lowered by

placing small beaker containing water inside the chamber. As the solvent

evaporated, instead of forming the coffee ring stain, LC phase was uniformly formed

on the grid (Figure 8.10).

Figure 8.10. a) A schematic showing the preparation of a gold NR assembly on a

TEM grid, b) A polarizing microscope image showing LC phase uniformly formed

over the grid during the evaporation.

The completely dried sample was observed with TEM. Inhomogenity of

local concentration of gold NR solution caused the different self assembly of NRs on

the substrate (Figure 8.11). At low number density, 2 to 4 NRs form aggregates and

exist sparsely in an isotropic state. As the number density increases, NRs start to

a) b)



201

form a monolayer of partially nematic-like phase. With further increase of number

density, NRs form smectic-like phase. For short NRs (aspect ratio is 3), the smectic-

like assembly showing both orientational and positional order was observed (Figure

8.12). The ordered assembly was observed for the very long NRs (aspect ratio is 14).

In this case, even though the suspension contains some of the spherical byproducts,

phase transition from isotropic to nematic and smectic phase can be confirmed by

the self assembly of dried suspension (Figure 8.13).

Figure 8.11. TEM images of gold NR assembly. Aspect ratio is 6. a) Isotropic phase

b) nematic-like assembly, and c) smectic-like assembly.

Figure 8.12. TEM images of gold NRs assembly. Aspect ratio is 3. a) Isotropic

phase and b) transition from isotropic to smectic-like assembly.

a) b) c)

200nm100nm100nm

a) b)



202


and b) nematic-like assembly.

The order parameter was measured to characterize the liquid

crystalline phase. The angle between the director and the long axis of

each rod was measured and the 2 dimensional order parameter,

θ= 2cosS was calculated for NRs of two different aspect ratios

(L/d=3 and 6) and plotted as a function of number density (

Figure 8.14).

Figure 8.14. Measured values of the gold NRs number density and the order

parameter, S. (a) The aspect ratio is 3 and (b) the aspect ratioj is 6. Number density

was calculated by counting the number of NRs per 10000nm2.

a) b)

a) b)

0

0.2

0.4

0.6

0.8

1

0.000 0.001 0.002 0.003

number density

o r d e r p a r a m e t e r

0

0.2

0.4

0.6

0.8

1

0.000 0.001 0.002 0.003

number density

o r d e r p a r a m e t e r



203

Bates el al. studied two-dimensional hard rod fluids consisting of

spherocylinders confined to lie in a plane using Monte Carlo simulations24. For long

rods, a 2D nematic phase is observed at high density. The transition from this phase

to the low density isotropic phase is continuous. For short rods the nematic phase

disappears so that, the solid phase undergoes a first order transition directly from

smectic to an isotropic phase. We found in our experiments that transition from

isotropic to higher order structure is continuous for longer NRs but for shorter NRs

isotropic phase undergoes a first order transition directly to an smectic phase. These

results are in good agreement with the simulations.

8.3.4. Patterns formed by LC in evaporating solution.

Figure 8.15 shows the deposit resulting from dried drops of different initial

volume fraction of gold NRs (aspect ratio 6) and evaporation conditions arranged left

to right in order of decreasing volume fraction of gold NRs. For each image, a drop

of 2µl gold NR solution was placed on the glass slide, which resulted in fluid films of

3mm in diameter on the substrate.

The upper row shows the images of dried deposit from the slow evaporation.

In this case, the substrate was kept in a closed environment. It takes 1hr for the drop



204

to be completely dried. We only observed the formation of single ring/band at the

contact line regardless of the volume fraction of the solute. The width of the band

decreases as the volume fraction of NRs decreases.

The bottom row shows the images of dried deposit from the fast evaporation.

In this case, the substrate was kept in open environment until the water evaporates

completely. The time to complete the evaporation is about 11mins. In (f) there is a

dark ring formed at the edge. As the volume fraction decreases, we observed

multiple rings formed interior of this ring while the width of the ring also slightly

decreased (g, h and i). Upon further decreasing concentration, very thin ring was

formed without additional rings inside (h).

The patterns formed by various concentrations can be visualized by polarized

optical microscopy because of the birefringence of ordered assembly of gold NRs,

and is shown in Figure 8.16.



205

Figure 8.15. “Coffee stain” formed by drying drops of gold NR solution. The images

in the upper row show the drying drops from slow evaporation. The volume fraction

decreases from left to right: a) 1×10-5 , b) 5×10-6 ,c) 3.3×10-6 d) 2.5×10-6 and f)

1.25×10-6. The images at the bottom row are from fast evaporation. The volume

fraction of the drops in the same column is identical. The scale bar is 200µm.

Figure 8.16. The patterns formed by various concentrations and different

evaporation condition were visualized by polarized optical microscopy for the

samples shown in Figure 8.15. Only the edge areas are illustrated. The scale bar is

200µm.

(a) (b) (c) (d) (e)

(f) (g) (h) (i) (j)

(a) (b) (c) (d) (e)

(f) (g) (h) (i) (j)



206

The multiple ring pattern formation has been observed when a suspension

droplet dries on a solid substrate25-28. It can be understood by a similar mechanism

to that of the “coffee stain” formation and closely related to the receding of contact

line. A suspension of latex particles has been usually chosen to investigate this

phenomenon. Adachi et al. 25 examined the drying process and found that the droplet

contact line shrinks toward the center of the droplet with an oscillatory motion

causing generation of particle array film at the contact line. Deegan26 further

studied the parameters affecting the patterns of dried drops and found that depending

on the concentration and the size of the drop, multiple rings with different sub-

structure were formed. Recently, this mechanism has been adapted to prepare

ordered structure of NPs and biomolecules27,28.

The particles assemble to form a ring at the droplet contact line due to the

convective flow induced by the evaporation of water from the filmsurface near the

contact line. In the case of gold NRs, the observation of the drop through polarized

microscopy gives the opportunity to examine the dynamic process of assembly due to

the formation of LC phase. While the birefringent contact line grows to a band with

the emerging LC droplets from the interior, the contact line shrinks toward the

center. The area where the contact line passes, birefringence becomes weaker.

The bright band seems to move inward because only the leading edge shows intense



207

birefringence. The relation between a change in the velocity of particles and that of

the shrinking contact line is the key to understand the various patterns observed.

Highest evaporation rate

Particles carried to the edge

Next pinning site

Contact line moving



a)

b)



Next pinning site

Contact line moving



a)

b)

Figure 8.17. a) A schematic shows the accumulation of gold NRs at the contact line in

the case of slow evaporation. Most of the particles can be carried to the edge forming

a single dark ring. b) In the case of fast evaporation, contact line starts to recede while

the particles move toward the edge resulting in the space between pinning sites.

When the evaporation is slower, the contact line is pinned long enough that a

large contact line deposit can form resulting in a single ring which contains most of

the particles in solution. And the width of the ring depends on the initial

concentration of gold NRs. When the evaporation is faster, the contact line starts to

shrink before enough amount of gold NRs accumulate at the outer edge leaving the

spaces between the rings. And the shinking contact line drags the particle toward to

the center. The contact line is arrested again when considerable amount of particles

accumulates and the rings chronicle the moments of arrested contact line motion

(Figure 8.17).



208

Figure 8.18. The well defined concentric multiple ring pattern formed by diluted gold

NR solution. The image is taken under optical microscope with cross polarizers. The

scale bar is 100 µm.

Figure 8.18 shows the well defined concentric multiple ring pattern that forms

on evaporation. The distance between the contact lines as well as the time between

pinning and sticking events do not show periodicity. Therefore the motion of the

contact line can not be described as a simple sinusoidal response. We believe that with

precise control of the drop size, roughness of the substrate and evaporation condition

(inadvertent liquid heating caused by droplet illumination), one might obtain better

data to conduct more systematic analysis of the multiple ring patterns.

8.4. Conclusions

We presented experimental results on the observation of lyotropic LC phase of

gold NRs. The dynamic phase transition was observed by evaporating drops of gold

NR solution on a glass slide. As the solution evaporates, the particles are carried

toward the contact line causing the local concentration near the contact line to become

higher than transition concentration. The evaporation rate and concentration of the



209

solute play an important role in determining not only the pattern of “coffee stain”

but also the multiple ring formation. We found that the pinning/depinning motion of

the contact line is responsible of this phenomena. But in our observation, the motion

of the contact line is essentially stochastic. We believe that the ability to deposit a

ring and multiple rings of gold NRs on a solid substrate provide potential for surface-

patterning applications.



210

References

1 Zocher, H. “Uber freiwillige Strukturbildung in Solen (Eine neue Art anisotrop

flqssiger Medien) ” Z. Anorg. Allg. Chem. 1925, 147 , 91.

2 Langmuir, I. “The role of attractive and repulsive forces in the formation of tactoids,

thixotropic gels, protein crystals and coacervates”, J. Chem. Phys. 1938, 6 , 873.

3 Bawden,F.C.; Piries,N.W.; Bernal, J.D.; Fankuchen, I. “Liquid crystalline

substances from virusinfected plants” Nature 1936, 138, 1051.

4 Buining, P. A.; Philipse, A. P.; Lekkerkerker, H. N.W. “Phase Behavior of Aqueous

Dispersions of Colloidal Boehmite Rods” Langmuir 1994, 10, 2106.

5 Maeda, H.; Maeda,Y. “An Atomic Force Microscopy Study of Surface Structures of

Colloidal α -FeOOH Particles Forming Smectic Layers” Nano Lett. 2002, 2, 1073

6 Vroege, G. J.; Lekkerkerker, H. N. W. “Phase transitions in lyotropic colloidal and

polymer liquid crystals” Rep. Prog. Phys. 1992, 55, 1241.

7 Onsager, L. “The effects of shape on the interaction of colloidal particles” Ann. N. Y.

Acad. Sci. 1949, 51, 627.

8 Bolhuis,P.; Frenkel,D. “Tracing the phase boundaries of hard spherocylinders” J.

Chem. Phys. 1997, 106 , 666.

9 Maeda, H.; Maeda,Y. “Liquid Crystal Formation in Suspensions of Hard Rodlike

Colloidal Particles: Direct Observation of Particle Arrangement and Self-Ordering

Behavior” Phys. Rev. Lett . 2003, 90, 018303.

10 Lamarque-Forget, S.; Pelletier,O.; Dozov, I.; Davidson,P.; Martinot-Lagarde,P.;

and Livage,J. “Electrooptic Effects in the Nematic and Isotropic Phases of

Aqueous V2O5 Suspensions” Adv. Mater . 2000, 12, 1267.

11 Lemaire, B.J.; Davidson, P.; Ferre, J.; Jamet, J.P.; Panine, P.; Dozov, I.; Jolivet, J.P.

“Outstanding magnetic properties of nematic suspensions of goethite (alpha-

FeOOH) nanorods” Phys. Rev. Lett . 2002, 88, 125507.

12 a) Kim, F.; Kwan, S.; Akana, J.; Yang, P. “Langmuir-Blodgett Nanorod Assembly”



212

22 Zasadzinski, J.A.N.; Meyer, R.B. “Molecular Imaging of Tobacco Mosaic Virus

Lyotropic Nematic Phases” Phys.Rev.Lett. 1986, 56 , 636.

23 The images of a,c and d were taken by Dr. H.J.Lee at Korea Research Institute ofStandards and Science using one of the gold NRs prepared in our study.

24 Bates, M.A.; Frenkel, D. “Phase behavior of two-dimensional hard rod fluids” J.Chem. Phys. 2000, 112, 10034.

25 Adachi,E.; Dimitrov,A.S.; Nagayama,K. “Slip stick Stripe Patterns Formed on a

Glass Surface during Droplet Evaporation” Langmuir 1995, 11, 1057.

26 Deegan,R.D., “Pattern formation in drying drops” Phys.Rev.E 2000, 61,

475.

27 Small,W.R.; Walton,C.D.; Loos, J.; Panhuis, M. “Carbon Nanotube Network

Formation from Evaporating Sessile Drops” J. Phys. Chem. B 2006, 110, 13029

28 Smalyukh,I.I.; Zribi,O.V.; Butler,J.C.; Lavrentovich,O.D.; Wong,J.C.L. “Structure

and Dynamics of Liquid Crystalline Pattern Formation in Drying Droplets of DNA”

Phys.Rev.Lett. 2006, 96 , 177801.



213

9 Chapter IX

Conclusions and Recommendations for the Future Work

9.1. Conclusions

The major findings of the work are summarized below:

We have shown that the morphology of gold nanorods(NRs) is controlled

by manuplating the reduction rate of gold ion and the adsorption kinetics of the

surfactant on the surface of gold nanoparticle. We prepared gold NRs with low

polydispersity and byproducts using improved seed mediated method. With

modifying the reduction steps, higher aspect ratio of gold NRs can be obtained.

We found that centrifugation can be the method of choice for the separation

when the synthesis results in particles with different sizes and shapes. Different

sizes and shapes of particles result in sufficiently different sedimentation coefficients

leading to efficient separation of particles. Thus, the efficient separation of gold

NRs from mixture of shapes was achieved by understanding the hydrodynamics of

NRs and nanospheres undergoing centrifugation. The purity of the NRs is higher

than 99% and the yield of separation through a round of centrifugation can be

achieved up to 78% by increasing CTAB concentration which affects the colloidal

interaction.



214

The optical properties of resulting refined gold nanorods were compared to the

predictions of existing theories, and the main parameters affecting them such as

aspect ratio and the dielectric properties of the environment were discussed.

Gold NRs with well defined aspect ratio were introduced into a PVA matrix

by means of solution-casting techniques. The film was drawn to induce the uniaxial

alignment of gold NRs to be used as color polarizing filters. We observed strong

optical dichroism in the visible as well as near – IR region which can be utilized to

make polarizers over a large range of wavelengths.

The various patterns of self assembly were formed by deposition of gold NR

solution on a TEM grid. The different patterns can be explained by drying mediated

self assembly where the coverage ratio and evaporation condition are among the

important factors to determine the characteristics of the pattern.

The phase transition to a liquid crystalline phase was observed by evaporating

the drop of gold NR solution on a glass microscope slide. As the solution evaporates,

the particles are carried toward the contact line causing the local concentration to

become higher than the concentration required for the phase transition. The

evaporation rate and concentration of the solute play an important role to result in not

only the pattern of the “coffee stain” but also the multiple ring formation. We



215

found that the pinning/depinning motion of the contact line is responsible for this

phenomenon.

9.2. Recommendations for future works.

In nanomaterial research, there is rapidly growing interest in the bulk

fabrication of hybrid nanostructures with discrete domains of different materials

arranged in a controlled fashion. The unique characteristics of such structures are the

ability to independently optimize the dimension and composition of the individual

components and thus providing entirely novel properties via the coupling between

components1.

Possible routes to fabricate the building blocks for these structures include the

synthesis of compositionally heterogeneous nanoparticles (usually by the growth of

the 2nd component on the surface of an as-made nanoparticle), bi-phase interfacial

modifications, or phase separation of bi-component mixtures on a nanoparticle

surface2,3. Optimally, theses building blocks are designed to naturally assemble into

the final structure through asymmetric particle-particle interaction. This assembly

can be further directed by providing predetermined “instructions” for assembly, such

as by modifying the nanoparticle surface with biomacromolecules4.

Considering the fact that the successful applications of nanoparticles require



216

the ability to tune their properties by controlling size, shape and surface modifications,

gold NPs are an excellent foundation to expand, given the substantial progress in the

synthesis of complex shapes in the past few years5.

We have demonstrated that different shapes of Au NPs besides nanorods can

be prepared by controlling the various parameters that affect the growth kinetics.

The unique surface plasmon resonance of these complex shapes endows many

interesting properties such as color tunablity, enhanced fluorescence, and surface-

enhanced Raman scattering. Combined, these properties make metal nanoparticles

valuable candidates for future nanoelectronics and nanophotonic applications once

appropriate techniques are developed to allow the controlled generation of artificial

assemblies.

We suggest combining several routes to fabricate hybrid NPs including the

generation of heterogeneous nanoparticles and the phase separation of bi-component

mixtures on a nanoparticle surface. Fabrication of hybrid NP will involve these

steps: (1) synthesis of gold NP with different morphologies (2) epitaxial growth of 2nd

component on the pre-made Au NPs to generate hybrid NPs, and (3) modification of the

hybrid NP surface utilizing pattern formation (phase separation) of multicomponent

oligomer/polymer on the surface of the particles. These structures will be utilized in

developing negative index materials and hybrid piezoelectric sensors.



References

1 Perro,A.;Reculusa,S.;Ravaine,S.;Bourgeat-Lami,E.;Duguet, E., “Design and

synthesis of Janus micro- and nanoparticles”, J.Mat.Chem. 2005, 15, 3745.

2 Roan, J. “Soft Nanopolyhedra as a Route to Multivalent Nanoparticles”

Phys.Rev.Lett . 2006, 96 , 248301.

3 Glogowski, E.; He,J.; Russell, T.P.; Emrick,T. “Mixed monolayer coverage on

park kyoungweon 200612 phd

Documents